:: ghana statistical service :: - reflect the uk governments official … · 2016. 11. 21. · this...

This study was funded by UK aid from the UK government. However, the views expressed do not necessarily

reflect the UK government’s official policies.

Suggested citation:

Institute of Development Studies (IDS), Ghana Statistical Services (GSS) and Associates (2016), Domestic

Violence in Ghana: Incidence, Attitudes, Determinants and Consequences, Brighton: IDS.

© Institute of Development Studies 2016

ISBN: 978-1-78118-314-4

This is an Open Access report distributed under the terms of the Creative Commons Attribution 4.0

International licence, which permits unrestricted use, distribution, and reproduction in any medium,

provided the original authors and source are credited. http://creativecommons.org/licenses/by/4.0/legalcode

http://creativecommons.org/licenses/by/4.0/legalcode

1

Contents

List of tables and figures 4

Foreword 7

Acknowledgements 9

Research team 12

Acronyms and abbreviations 14

Executive summary 15

1. Background to the study 20

1.1. Domestic violence in Ghana 21

1.2. The study 25

2. Conceptual framework 27

2.1. Types and perpetration of domestic violence 31

2.2. Attitudes towards domestic violence 34

2.3. Determinants of domestic violence 35

2.4. Consequences of domestic violence 39

2.5. Use of and access to victims’ support services and institutions 41

3. Methodology 44

3.1. Quantitative instruments 44

3.2. Qualitative instruments 46

3.3. Sampling design: quantitative survey 48

3.4. Sampling design: qualitative fieldwork 49

3.5. Training and pre-testing 50

3.6. Fieldwork implementation 51

3.7. Response rates 52

3.8. Data analysis 57

3.9. Ethical considerations 58

4. Empirical results 61

4.1. Incidence of violence (domestic and non-domestic) against women and men in Ghana 61

4.1.1. Social violence 62

4.1.2. Physical violence 73

4.1.3. Sexual violence 83

4.1.4. Psychological violence 93

4.1.5. Economic violence 103

4.2. Incidence of domestic violence against women and men in Ghana 113

4.2.1. Domestic social violence 114

4.2.2. Domestic physical violence 121

4.2.3. Domestic sexual violence 128

4.2.4. Domestic psychological violence 135

4.2.5. Domestic economic violence 142

4.2.6. Summary 149

2

4.3. Perpetration of domestic violence in Ghana 150

4.3.1. Violence perpetration within domestic relations 150

4.3.2. Profile of domestic violence perpetrators 153

4.3.3. Domestic violence against children 158

4.4. Attitudes towards domestic violence in Ghana 159

4.4.1. Attitudes towards domestic social violence 160

4.4.2. Attitudes towards domestic physical violence 166

4.4.3. Attitudes towards domestic sexual violence 179

4.4.4. Attitudes towards domestic psychological violence 182

4.4.5. Attitudes towards domestic economic violence 184

4.5. Determinants of domestic violence in Ghana 185

4.5.1. Age and intergenerational effects 187

4.5.2. Socio-economic effects 189

4.5.3. Marital status 193

4.5.4. Geographical effects 196

4.5.5. Patriarchal and gender norms 197

4.5.6. Exposure to other forms of violence 200

4.6. Consequences of domestic violence in Ghana 201

4.6.1. Physical health effects 202

4.6.2. Mental health and emotional effects 206

4.6.3. Effects on work and daily activities 208

4.6.4. Consequences for children 209


4.7.1. Seeking external help 211

4.7.2. Public services and institutions 215

4.7.3. Knowledge about support services and institutions 219

4.7.4. Levels of satisfaction about support services and institutions 223

5. Summary and discussion 225

5.1. Incidence of domestic violence in Ghana 225

5.2. Attitudes towards domestic violence in Ghana 226

5.3. Determinants of domestic violence in Ghana 227

5.4. Consequences of domestic violence in Ghana 230


6. Way forward and policy recommendations 233

6.1. Recommendations with respect to the incidence of domestic violence in Ghana 234

6.2. Recommendations with respect to attitudes towards domestic violence in Ghana 236

6.3. Recommendations with respect to the determinants of domestic violence in Ghana 238

6.4. Recommendations with respect to the consequences of domestic violence in Ghana 240

6.5. Recommendations with respect to the use of and access to victims’ support services and

institutions in Ghana 241

6.6. Recommendations for future research on domestic violence in Ghana 244

References 247

3

Appendices 259

Appendix A: Incidence of domestic violence worldwide 260

Appendix B: Timeline of major policies related to domestic and gender-based violence 263

Appendix C: Summary of qualitative sample, methods and instruments 265

Appendix D: Distribution of enumeration areas across regions and survey weights 266

Appendix E: Regression tables 269

Questionnaire annex 275

4

List of tables and figures

Table 1: Response rates to household survey (in percentages) .............................................. 53

Table 2: Response rates to the individual survey: women (in percentages) ........................ 54

Table 3: Response rates to the individual survey: men (in percentages) .............................. 54

Table 4: Distribution of survey respondents by key socio-economic characteristics

(weighted) ...................................................................................................................................... 55

Table 5: Incidence of violence in Ghana 2015 (in percentages) .............................................. 62

Table 6: Percentage of women who have experienced domestic or non-domestic social

violence over their lifetime .......................................................................................................... 65

Table 7: Percentage of women who experienced domestic or non-domestic social violence

in the last 12 months ..................................................................................................................... 67

Table 8: Percentage of men who have experienced domestic or non-domestic social


Table 9: Percentage of men who experienced domestic or non-domestic social violence in

the last 12 months .......................................................................................................................... 71

Table 10: Percentage of women who have experienced domestic or non-domestic physical


Table 11: Percentage of women who experienced domestic or non-domestic physical

violence in the last 12 months ..................................................................................................... 77

Table 12: Percentage of men who have experienced domestic or non-domestic physical


Table 13: Percentage of men who experienced domestic or non-domestic physical


Table 14: Percentage of women who have experienced domestic or non-domestic sexual


Table 15: Percentage of women who experienced domestic or non-domestic sexual


Table 16: Percentage of men who have experienced domestic or non-domestic sexual


Table 17: Percentage of men who experienced domestic or non-domestic sexual violence

in the last 12 months ..................................................................................................................... 91

Table 18: Percentage of women who have experienced domestic or non-domestic

psychological violence over their lifetime ................................................................................. 95

Table 19: Percentage of women who experienced domestic or non-domestic psychological


Table 20: Percentage of men who have experienced domestic or non-domestic

psychological violence over their lifetime ................................................................................. 99

Table 21: Percentage of men who experienced domestic or non-domestic psychological

violence in the last 12 months ................................................................................................... 101

Table 22: Percentage of women who have experienced domestic or non-domestic

economic violence over their lifetime ...................................................................................... 105

Table 23: Percentage of women who experienced domestic or non-domestic economic


5

Table 24: Percentage of men who have experienced domestic or non-domestic economic

violence over their lifetime ........................................................................................................ 109

Table 25: Percentage of men who experienced domestic or non-domestic economic


Table 26: Incidence of domestic violence in Ghana 2015 in the 12 months prior to the

survey (in percentages) .............................................................................................................. 113

Table 27: Percentage of domestic social violence over the last 12 months ......................... 114

Table 28: Percentage of women who experienced domestic social violence over the last

12 months ..................................................................................................................................... 117

Table 29: Percentage of men who experienced domestic social violence over the last 12

months .......................................................................................................................................... 119

Table 30: Percentage of domestic physical violence over the last 12 months..................... 121

Table 31: Percentage of women who experienced domestic physical violence over the

last 12 months .............................................................................................................................. 124

Table 32: Percentage of men who experienced domestic physical violence over the last

12 months ..................................................................................................................................... 126

Table 33: Percentage of domestic sexual violence over the last 12 months ........................ 128

Table 34: Percentage of women who experienced domestic sexual violence over the last

12 months ..................................................................................................................................... 131

Table 35: Percentage of men who experienced domestic sexual violence over the last 12

months .......................................................................................................................................... 133

Table 36: Percentage of domestic psychological violence over the last 12 months ........... 135

Table 37: Percentage of women who experienced domestic psychological violence over

the last 12 months ........................................................................................................................ 138

Table 38: Percentage of men who experienced domestic psychological violence over the

last 12 months .............................................................................................................................. 140

Table 39: Percentage of domestic economic violence over the last 12 months .................. 142

Table 40: Percentage of women who experienced domestic economic violence over the

last 12 months .............................................................................................................................. 145

Table 41: Percentage of men who experienced domestic economic violence over the last

12 months ..................................................................................................................................... 147

Table 42: Socio-economic groups with highest incidence of domestic violence ................ 149

Table 43: Perpetrators of domestic violence as reported by women (in percentages) ...... 151

Table 44: Perpetrators of domestic violence as reported by men (in percentages) ........... 152

Table 45: Percentage and number of domestic violence perpetrators ................................. 153

Table 46: Number and percentage of perpetrators reporting violence against same or

another sex ................................................................................................................................... 154

Table 47: Percentage of women reporting perpetration of domestic violence ................... 156

Table 48: Percentage of men reporting perpetration of domestic violence ........................ 157

Table 49: Perpetration of domestic violence against individuals who reported domestic

violence and lived with their parents (in percentages) ......................................................... 158

Table 50: Responses to first social violence scenario (in percentages) ................................ 161

Table 51: Responses to second social violence scenario (in percentages) ........................... 163

Table 52: Responses to first physical violence scenario (in percentages) ........................... 167

6

Table 53: Responses to second physical violence scenario (in percentages) ...................... 169

Table 54: Percentage of respondents agreeing with wife-beating if the wife… ................. 172

Table 55: Percentage of respondents who think that a man has a good reason to hit his

wife if: ........................................................................................................................................... 173

Table 56: Percentage of respondents stating that wife-beating is acceptable if the

woman… ...................................................................................................................................... 175

Table 57: Views about statement ‘To bring up, raise or educate a child properly, the child

sometimes needs to be physically punished’ (in percentages) ............................................. 178

Table 58: Views about statement ‘If a woman wears revealing clothes, it is okay for men

to make some sexual comments as she walks in her neighbourhood’ (in percentages) .. 180

Table 59: Views about statement ‘If a woman wears revealing clothes, it is her fault if she

is raped’ (in percentages) ........................................................................................................... 181

Table 60: Regions significantly associated with increased experience of domestic violence

(by type of domestic violence and sex) .................................................................................... 197

Table 61: Relationship between experience of domestic violence and physical health

(women) ........................................................................................................................................ 204


(men) ............................................................................................................................................. 205

Table 63: Percentage of respondents who have experienced domestic violence reporting

serious mental illness (by type of domestic violence and sex) ............................................. 206

Table 64: Effect of domestic violence on men and women’s mental health ....................... 207

Table 65: Percentage of respondents who experienced domestic violence who reported

seeking help (by sex) ................................................................................................................... 212

Table 66: Reasons for approaching external help after exposure to any form of domestic

violence by sex (in percentages) ................................................................................................ 212

Table 67: Reasons for not seeking external help after exposure to any form of domestic

violence by sex (in percentages) ................................................................................................ 213

Table 68: First choice of help (in percentages) ........................................................................ 214

Table 69: Respondents’ knowledge about public services (in percentages) ....................... 221

Table 70: Satisfaction levels with public services (number of respondents) ...................... 223

7

Foreword

Domestic violence, in the form of physical or sexual violence, is experienced by 35 per cent

of women globally during their lifetime. Emotional abuse and controlling behaviour is

experienced by 75 per cent of women across the world. It is also acknowledged that men

experience domestic violence. Nationally representative statistics on the incidence of

domestic violence are, however, hard to come by.

This study represents a comprehensive attempt to estimate the incidence of domestic

violence in Ghana, and analyse the attitudes that sustain domestic violence, its

determinants and its consequences. The study estimates that 28 per cent of women and 20

per cent of men experienced domestic violence in 2015.

Domestic violence causes immediate devastating consequences to those affected: physical

injuries, mental health problems and poor well-being, to name but a few. But it also has

long-term, far-reaching effects, including persistent inequalities between men and women,

which limit women and girls’ abilities to fulfil their potential. Those who experience

domestic violence have longer-term poor physical and mental health, higher risk of

contracting HIV and sexually transmitted diseases, and restricted choices in terms of

accessing education and jobs. It is also likely that they will have lower levels of education,

income and productivity. These consequences are not only life-changing for the women

and men concerned, but also for their families, communities and wider societies, as gender

inequalities become entrenched.

This innovative mixed-methods study, commissioned by the Ministry of Gender, Children

and Social Protection in Ghana and the UK Department for International Development,

and funded by UK aid, provides an in-depth understanding of the incidence, attitudes,

causes and consequences of domestic violence in Ghana, as well as investigating the

effectiveness of existing institutional support offered to women and men.

Ghana has, in many ways, led the way with its pioneering legislation on domestic violence,

which includes various forms of economic abuse, acknowledges that perpetrators and

survivors do not have to be married or related by blood ties, and takes a broader

perspective on access to justice than other countries, allowing for mediation by alternative

dispute resolution methods.

The Domestic Violence Act 732, adopted by Parliament in 2007, outlines a comprehensive

legal framework for the prevention of and protection against domestic violence and

criminalises various acts of physical and sexual violence, economic and psychological

abuse, and intimidation in domestic relations.

Despite this innovative legislative work, the results of this study show that there is more

work to be done to ensure that the laws are enacted and understood by the broader

population. The findings of this work will be used to formulate ongoing policies and

8

programmes to ensure that, moving forward, fewer and fewer people will be affected by

domestic violence in Ghana.

Hon. Nana Oye Lithur

Minister for Gender, Children and Social Protection

9

Acknowledgements

This report discusses the main findings of a comprehensive study of the incidence,

determinants and consequences of domestic violence, attitudes towards domestic violence

and the effectiveness of services that support women and men who experience domestic

violence in Ghana. The study was commissioned by the Domestic Violence Secretariat of

the Ministry of Gender, Children and Social Protection (MoGCSP) of the Government of

Ghana and the UK Department for International Development (DFID), and funded by UK

aid.

The study was led by the Institute of Development Studies (IDS) in the United Kingdom

(UK) and carried out in cooperation with the Ministry of Gender, Children and Social

Protection, the Ghana Statistical Service (GSS) and researchers at the Institute of Statistical,

Social and Economic Research (ISSER) in Accra.

The research team is very grateful to Professor Akosua Adomako Ampofo (Institute of

African Studies at the University of Ghana), who reviewed the quantitative and qualitative

instruments and participated in the quantitative pilot training exercises; to Mrs Dorcas

Coker-Appiah for providing invaluable guidance during training to the enumerating team

about domestic violence and ways of relating to potential victims of domestic violence

during the survey; and to Mr Adolf Awuku Bekoe (Coordinator, DV Coalition) for his help

in coordinating referral support provided by different organisations to the study

participants who requested it across the regions covered by the study. Mr Adolf Awuku

Bekoe also provided invaluable psychological support to the fieldwork team when needed

via a telephone helpline set up for the study.

At the MoGCSP, the study was supported by the Honourable Minister Nana Oye Lithur

and her team. The Honourable Minister chaired the study’s Steering Committee, which

provided feedback on the study’s outputs at key stages, facilitated contacts with key

stakeholders beyond the Steering Committee representatives, and supported the outreach

and uptake of the study.

The Steering Committee was comprised of Mr Kwesi Armo-Himbson (Chief Director,

MoGCSP), Mrs Efua A. Anyanful (MoGCSP), Ms Fauzia Issaka (DFID Ghana), Ms Lynne

Henderson (DFID Ghana), Mrs Patience Opoku (MoGCSP), Mrs Abena Annobea Asare

(MoGCSP), Mr Kwabena Twumasi (Human Trafficking Secretariat), Mrs Lilian Ayete-

Nyampong (Commission on Human Rights and Administrative Justice – CHRAJ), ACP

Habiba Twumasi-Sarpong (Domestic Violence Victim Support Unit – DOVVSU), Mr Adolf

Awuku Bekoe (DV Coalition), Mrs Ruth Aba Grant (Network for Women’s Rights –

NETRIGHT), Professor Felix A. Asante (ISSER), Mr Daniel Degbotse (Ministry of Health)

and Mr Clarke Noyoru (MoGCSP). The research team is grateful to Mrs Victoria Natsu for

the logistical support she provided to meetings and interactions with the Steering

Committee throughout the study.

10

The study benefited from excellent input, discussions and support from DFID Ghana. Ms

Lynne Henderson provided support to the design of the study, participated in several

discussion and reviewed drafts of the report. Mr Archie Laing provided invaluable

logistical support to the team throughout the study. Ms Fauzia Issaka engaged with the

research team at all stages of the study, provided comments on several drafts of the

fieldwork instruments, participated in numerous research discussions, helped the team to

identify and liaise with key stakeholders and reviewed in detail multiple drafts of the

report. The depth and quality of her feedback has helped substantially to improve the

report. The research team is extremely grateful for her unfaltering support and

encouragement during the entire study.

At IDS, the research team benefited from fantastic research assistance by Marco Carreras

and Eric Casper. Deborah West and Veronica Moore managed the contractual and financial

aspects of the study. Deborah West also provided invaluable assistance with

communication activities and editorial support. The study also benefited from excellent

comments and support from Dr Jerker Edstrom, Professor John Gaventa, Professor Melissa

Leach and, especially, Dr Mariz Tadros, who read and commented on the entire initial draft

of the report.

The study’s qualitative and quantitative instruments and various drafts of the report were

reviewed by four members of the study’s Advisory Board: Mrs Dorcas Coker-Appiah

(Executive Director of the Gender Centre, Accra), Dr Akosua K. Darkwah (Director, Centre

for Gender Studies and Advocacy, University of Ghana), Dr Nata Duvvury (Co-Director

of the Global Women’s Studies Programme, National University of Ireland, Galway) and

Dr Katherine Brickell (Royal Holloway University of London). The research team is very

grateful for their contribution, support and invaluable and constructive comments at all

stages of the study.

The research team would also like to thank key stakeholders from the Gender Centre, DV

Coalition, DV Secretariat, Ark Foundation, International Federation of Women Lawyers

(FIDA), Leadership and Advocacy for Women in Africa (LAWA) and other organisations,

in particular Ms Dinah Adiko (consultant, DV Secretariat, MoGCSP), Ms Juliana

Amponsah (former Acting Executive Secretary, DV Secretariat, MoGCSP), Superintendent

Laurencia Akorli (Deputy Coordinating Director of DOVVSU), Mrs Barbara Ayesu

(Coordinator, LAWA (Ghana) Alumnae Inc.), Dr James Clayman (Medical Superintendent

and Obstetrician Gynaecologist at the Ga West Municipal Hospital), Mrs Dorcas Coker-

Appiah (Executive Director, Gender Center), Mrs Esther Darko-Mensah (Gender Centre),

Mrs Angela Dwamena-Aboagye (Executive Director, Ark Foundation), Dr Elvis Sadongo

(Head of Social Services Department at Ghana Police Hospital) and Mrs Jane Quaye

(Executive Director, FIDA-Ghana) for their warm welcome, encouragement and invaluable

support.

The research team is thankful for the outstanding work of the fieldwork teams at the Ghana

Statistical Service and the Institute of Statistical, Social and Economic Research.

11

Most importantly, the research team is extremely grateful to all the women and men who

took time from their daily activities to talk about their experiences – some quite painful

and distressing. This study would have never been possible without them.

12

Research team

Core research team: Ms Alia Aghajanian (IDS), Mr Anthony Amuzu (GSS), Dr Elizabeth

Asante (ISSER), Professor Patricia Justino (IDS) (Team Leader), Dr Marinella Leone (IDS),

Dr Elizabeth Mills (IDS), Ms Sheila Minkah-Premo (Apex Lawconsult), Ms Rebecca

Mitchell (IDS), Dr Catherine Müller (IDS), Dr Philomena Nyarko (Government Statistician,

GSS), Dr Pauline Oosterhoff (IDS) and Dr Jean-Pierre Tranchant (IDS).

Fieldwork team at the Institute of Statistical, Social and Economic Research: Mohammed

Ben Abdallah, Adom K.M. Acheampong (Supervisor), Otuo Acheampong, James Allotey,

Aboagye Amankona, Dinah Badmus, Daniel Danso, George Yaw Dugbatey, Daniel Gyasi,

Samuel Gyasi-Adonteng (Supervisor), Michael Johnson, Randy Korley, Eric Obeng, Joseph

K. Okrah and Emmanuel Nuetey Siakwa (Supervisor).

Fieldwork team at the Ghana Statistical Service: Dr Philomena Nyarko (Government

Statistician/Project Director), Mr Baah Wadieh (Deputy Government Statistician/Deputy

Project Director), Mr Anthony Amuzu (Acting Deputy Government Statistician/Survey

Director), Mr Peter Takyi Peprah (Project Coordinator), Mr Emmanuel George Ossei

(Trainer/Field Monitor), Mr Emmanuel Boateng (Trainer/Field Monitor), Mr Godwin Odei

Gyebi (Trainer/Field Monitor), Mr K.B. Danso-Manu (Field Monitor), Mrs Araba Forson

(Field Monitor), Mr Owusu Kagya (Field Monitor), Mrs Betty-Love Hermans Cobbinah

(Internal Auditor), Mr Robert Kwami (Director of Finance), Mr Kwadwo Asante-Mensah

(Director of Administration), Mrs Abena Asamoabea Osei-Akoto (Head, Data Processing),

Mrs Lucy Twumwaa Afriyie (Head, Sampling Design), Ms Hannah Konadu Frimpong

(Project Secretary), Mrs Rosalind Quartey (Head, Geographic Information System) and Ms

Elizabeth Arhin Donkor (Project Accountant).

The team of regional statisticians included Isaac Addae, Salifu Amadu, Chris Amewu,

Kobina Abeka Ansah, Chris Assem, Samuel Brefo Darkwa, Sixtus Dery, Festus Manuh,

Amatus S. Nobabuma and Ernest Nyarku. The team of field supervisors included Barnas

Abugre, Michael Opoku Acheampong, Jane Acolatse, Samuel Owusu Agyemang,

Dieudonnee Ankamah, Daniel Ansah, Emmanuel Atanga, Solomon Owusu Bempah,

Victor Owusu Boateng, Patrick Kwame Djangba, Felix Geli, William Ofosu, Stephen Tabi,

Joseph, Asiedu Tenkorang and Richard Boakye Yiadom.

The team of interviewers included Daniel Amos Abanyie, Nana Afua Acheampong, Mavis

Adebona, Dennis Addo, Judith Addo, John Agbadza, Michael Agyarkwa, Frank Agyei,

Habib Alhassan, Sandra Amartey, Anastasia Ammisah, Esther Adwoa Amoah, Jolene

Mantey Amoo, Caroline Anipah, Anthony Ankomah, Umul Klusum Arimiyaw, Konlan

Alhasssan Awal, Gloria Bekoe, Vida Akuamoah Boateng, Gladys Ataa Dabison, Mabel

Diameh-Bentil, Getrude Elleammoh, Afu Anita Elorm, William Eyimah, Bright Gyamfi,

Sulemana Idriss, Felicity Kuwornu, Emmanuel Kyei, Evelyn Kyeremateng, Mark Ato

Larbi, Mary Larbi, Zainab Ibrahim Maltiti, Happy Krah Mawuse, Richard Mills, Linda

Ntiamoah, Israel Kofi Menash Nyarko, Pricilla Opoku, Seth Opoku, Mercy Naa Quarshie,

13

Phylicia Saforo, Rabiatu Seidu, Edward Tebiesinaa, Linda Ama Tenkorang, Isaac Kwame

Yeboah and Benjamin Kusi Yentumi.

The field teams were supported by a team of drivers: Ebenezer Abrokwah, Emmanuel

Acquah, Isaac Addae, Joseph Afedzie, Elvis Agyei, David Agyemang, William

Akamaboro, Joseph Ampomah, George Ansah, Eugene Aniapam, Peter Ayensu, Samuel

Ayikwei, Emmanuel Chartey, Samuel Dam, Ebenezer Dinsey, Kwesi Donkor, Tweneboah

Gyesi, Jacob Jabotir, George Karikari, Simon K. Kasakwa, Wisdom Kuagbelah, James Nii

Laryea, Victor Mensah, Gershon Nornyibey, Emmanuel Oduro, Isaac Opare, Felix Osei,

Stephen Otoo and Stephen Sanbonaa.

14

Acronyms and abbreviations

ACHPR African Charter on Human and Peoples’ Rights

CEDAW UN Convention on the Elimination of All Forms of Discrimination

against Women

CHRAJ Commission on Human Rights and Administrative Justice

CSO Civil Society Organisation

DE Design effect

DEVAW Declaration on the Elimination of Violence against Women

DFID UK Department for International Development

DHS Demographic and Health Survey

DOVVSU Domestic Violence Victim Support Unit

DV Domestic violence

DV Secretariat Domestic Violence Secretariat

EA Enumeration area

FGD Focus Group Discussion

FGM Female Genital Mutilation

FIDA International Federation of Women Lawyers

Gender Centre Gender Studies and Human Rights Documentation Centre

GFLHS Ghana Family Life and Health Survey

GSS Ghana Statistical Service

GVSSN Gender Violence Survivors’ Support Network

HIV Human Immunodeficiency Virus

IDS Institute of Development Studies

IPV Intimate Partner Violence

JHS Junior High School

JSS Junior Secondary School

LAWA Leadership and Advocacy for Women in Africa

MICS Multiple Indicator Cluster Survey

MoGCSP Ministry of Gender, Children and Social Protection

NETRIGHT Network for Women’s Rights

NGO Non-governmental Organisation

NPPOA National Policy and Plan of Action

SDGs Sustainable Development Goals

SMI Serious mental illness

UK United Kingdom

UN United Nations

UNICEF United Nations Children’s Fund

USA United States of America

VAWG Violence against women and girls

WAJU Women and Juvenile Unit of the Ghana Police

WHO World Health Organization

WISE Women’s Initiative for Self-Empowerment

15

Executive summary

Domestic violence is a violation of basic human rights and a global policy concern.

Following several years of advocacy efforts by civil society and women’s rights

organisations, the Government of Ghana enacted the DV Act (act 732) in February 2007,

followed by the formulation of the National Policy and Plan of Action (NPPOA) in 2008.

The NPPOA laid out a strategy for the implementation of the Domestic Violence Act, under

the direction of the Domestic Violence Secretariat, part of the Ministry of Gender, Children

and Social Protection (MoGCSP). This study forms part of this strategy.

The overarching aim of the study is to provide an in-depth understanding of the attitudes

to and incidence, determinants and consequences of domestic violence in Ghana. The

study addresses the following questions:

1. What types and levels of violence are experienced by women, men, girls and boys

in Ghana, and who perpetrates them? What proportion of this violence is domestic

violence?

2. What are society, community and people’s attitudes towards domestic violence, and

how do attitudes and norms influence the acceptance of different forms of domestic

violence in Ghana?

3. What are the main determinants of domestic violence against women, men, girls

and boys at the individual, family, community and wider societal levels?

4. What are the consequences of domestic violence in Ghana for people’s physical and

mental health and their daily lives?

5. What institutional structures support victims of domestic violence, and what

proportion of victims are able to effectively access available services?

6. How can policies and interventions be (re)designed to better support victims of

domestic violence and prevent domestic violence in Ghana?

Conceptual approach. The conceptual framework adopted in the study – the critical social

ecological model – uses an explicit political economy lens to analyse the complex interplay

between the determinants and consequences of domestic violence at multiple levels of

analysis: the individual, the family, the community and the wider social and political

system. This conceptual framework formed the basis for the collection of new empirical

evidence across Ghana on: (i) the incidence and types of domestic violence; (ii) the attitudes

and norms that drive domestic violence; (iii) the determinants of domestic violence; (iv)

the consequences of domestic violence; and (v) the use of and access to victims’ support

services and institutions.

Conforming to international definitions and to definitions included in the 2007 Domestic

Violence Act (Act 732), the study analysed the incidence of the following forms of violence:

Social violence: defined as acts of controlling behaviour, such as preventing someone

from seeing friends or family; stopping someone from leaving the house; requiring

to know where someone is at all times; stalking; spreading false information, videos

16

or photos without permission; or forcing women to have an abortion.

Physical violence: defined as slapping, pushing, shoving, hitting, kicking, dragging

or throwing objects at someone; choking, strangling or burning someone; using a

weapon, hazardous chemicals or substances against someone; or kicking or pulling

someone’s external genitalia (for male respondents only).

Sexual violence: defined as acts of unwanted sexual comments or physical contact;

rape by physical force, or otherwise forced sex (for instance, by blackmail or threats);

denial of using protection during sex; a sexual partner hiding their HIV status;

sexual acts and intercourse that were performed on the basis of feeling there was no

option; or penetration with an object against someone’s will.

Psychological violence: defined as the use of insults, belittling or humiliation in private

or in front of others; threats of abandonment; being ignored or treated indifferently;

intimidations and acts aimed at scaring someone; threats of using weapons against

someone; or threats of hurting someone or someone one cares about.

Economic violence: defined as the denial of household money for expenses (chop

money) even if enough financial means are available; unsolicited taking of money;

control of belongings and spending decisions; damage to or destruction of

someone’s property; denial of the right to work; forcing someone to work against

their will; or denial of food and other basic needs.

Methodological approach. The study adopted a mixed-methods approach. This involved

the collection of primary quantitative data through a representative household-level

survey of 4,995 individuals, and the collection of qualitative in-depth data on key factors

that may explain the incidence of domestic violence in Ghana.

The main quantitative tool used in the study was a survey entitled the “Ghana Family Life

and Health Survey 2015” (GFLHS 2015). The GFLHS 2015 was conducted between April

and August 2015. Given the sensitive nature of the topic, the preparation of the survey took

six months and involved intense testing, piloting and training. This careful approach

ensured high response rates, ranging between 97.6 per cent in the Western region and 99.6

per cent in the Brong Ahafo region.

The GFLHS 2015 included one household and one individual questionnaire. The study

focused on men, women, boys and girls between the ages of 15 and 60 years. This age

group extends the usual age group included in international surveys (typically 15 to 49

years), to capture the incidence of domestic violence among older women and men.

Children below the age of 15 years old were outside the scope of the study. The GFLHS

2015 asked detailed questions about individual experiences of any of the five broad

categories of domestic violence that occurred in the 12 months prior to the survey.

The sensitivity of the research topic raised several ethical and methodological challenges

that were carefully addressed during the study. The safety of respondents and the research

team was paramount, and guided all decisions in the study. Protecting the confidentiality

of the respondents was essential to ensure their safety and data quality, and all researchers

17

were carefully selected and received specialised training and ongoing support. Field

researchers were also trained to refer individuals requesting assistance to available local

services and sources of support.

Incidence of domestic violence. The study provides an in-depth analysis of the incidence,

attitudes, determinants and consequences of domestic violence in Ghana. The main results

show that 27.7 per cent of women and 20 per cent of men in Ghana have experienced at

least one type of domestic violence in the 12 months prior to the survey.

The most common form of domestic violence reported by women in the 12 months prior

to the survey was economic violence (12.8 per cent), followed by social violence (11.6 per

cent), psychological violence (9.3 per cent), physical violence (6.0 per cent) and sexual

violence (2.5 per cent).

The incidence of domestic violence was generally lower among men than among women.

The most common form of domestic violence experienced by men was psychological

violence (7.9 per cent), followed by social violence (7.7 per cent), economic violence (7.3

per cent), physical violence (2.1 per cent) and sexual violence (1.4 per cent).

The GFLHS 2015 is not comparable to prior surveys of domestic violence in Ghana because

it has collected data on broader and more in-depth definitions of domestic violence than

other surveys. It is, however, possible to compare the incidence of physical domestic

violence experienced by women and men in the 2008 DHS and in the 2015 GFLHS. The

results show, when using comparable categories of domestic physical violence, that

domestic physical violence among women decreased from 17.2 per cent in 2008 to 10.3 per

cent in 2015. Domestic physical violence among men decreased from 12.7 per cent in 2008

to 11.2 per cent in 2015.

Attitudes towards domestic violence. Respondents did not consider some of the

definitions of domestic violence used in the study (and in Act 732) to be a form of violence.

Many of the acts mentioned in the study were viewed as accepted social norms, defence

mechanisms or ways of resolving disputes. Only physical violence between partners was

consistently perceived as violence. Psychological violence, in particular, was generally

perceived as a means of defence and a coping strategy – for women in particular – to ward

off physical violence or to claim some autonomy or agency in contexts where they may not

have the ability to exert much power.

Women were on average more likely than men to condone forms of domestic violence: 23.1

per cent of women and 13.8 per cent of men found wife-beating acceptable, while 65.3 per

cent of women and 56.2 per cent of men fully agreed that women were to blame for rape if

they wore revealing clothes. These findings suggest that domestic violence in Ghana may

persist due to harmful social norms that hold women responsible for the violence they

experience.

18

Determinants of domestic violence. Age was found to be one of the key determinants of

domestic violence in Ghana. Young women and men (15–19 years old) were substantially

more likely to have experienced domestic violence than other age groups. Women aged

15–19 years were between 1.3 and four times more likely to experience any form of

domestic violence than, for instance, women aged 30–39 years. Differences in the incidence

of domestic violence across age groups were similar for men, with the exception of sexual

violence, which was more prevalent among men aged 20–24 years.

Exposure to violence in childhood was found to be strongly related to the likelihood of an

individual being a victim or perpetrator of violence in adulthood. Women who witnessed

any form of domestic violence as children were between 1.3 and three times more likely to

report domestic violence than women who did not witness violence during their

childhood. Men who witnessed social, psychological or economic violence in their homes

as children were between 1.2 and 2.2 times more likely to have experienced these forms of

violence in adulthood than other men.

Taken together, these findings suggest that Ghana may experience vicious

intergenerational cycles of domestic violence that need to be broken if further progress is

to be made in reducing domestic violence.

In addition to age, socio-economic factors were found to be among the most important

determinants of domestic violence in Ghana, albeit in complex ways:

Low asset levels were found to be strong determinants of psychological and

economic violence among women, but high asset levels were associated with higher

levels of social and sexual violence among men.

Tensions and fights over money or property were found to be leading determinants

of all forms of domestic violence across all regions.

Women with no education were generally (between 0.5 and 2.5 times) less likely to

experience domestic violence than women with some level of education. More

educated men were around three times more likely to experience economic or

sexual violence than less educated men. This result may be due to the fact that

women and men with more education may also be more likely to report domestic

violence, as emphasised in previous studies conducted internationally and in

Ghana.

The qualitative analysis suggested that factors such as poverty, unemployment and

economic shocks were identified as important determinants of domestic violence.

Overall, the determinants of domestic violence varied considerably among types of

violence, emphasising the need for domestic violence legislation in Ghana to continue to

distinguish between different forms of violence (as in Act 732).

Consequences of domestic violence. The GFLHS 2015 and the qualitative data revealed

serious adverse consequences of domestic violence on physical health, including injuries

and illness. The effects of domestic violence on physical health were observed almost

19

exclusively among women: 43.8 per cent of women who had been physically assaulted

declared having been ill in the 30 days prior to the survey – a rate that was one third higher

than among women who had not been physically assaulted (31.2 per cent). 42.3 per cent of

women who had experienced psychological violence reported ill health, compared to 30.9

per cent of women who had not experienced this form of domestic violence.

The results also showed a strong correlation between exposure to domestic violence and

mental health, particularly among women: women who reported serious mental illness

were on average around 25 per cent more likely to have experienced at least one form of

domestic violence than women who did not report serious mental illness. The association

between serious mental illness and domestic violence was generally lower among men.

Only physical violence was associated with men’s mental health status.

Domestic violence was also found to have adverse consequences on people’s daily life in

terms of their ability to work, go to school or do domestic work, their ability to concentrate

on daily activities, levels of confidence and feelings of living in fear, although few

respondents answered this question. The effects were larger for women than for men, and

for physical violence than for other types of violence.

Use of and access to services. The study investigated in detail how victims of domestic

violence make use of and access support services and institutions across Ghana. The results

showed that just over one in every three women and men who experienced domestic

violence approached a person, formal or informal organisation for help or support. The

majority of respondents who sought help first approached family members or a friend, but

14.6 per cent of men and 9 per cent of women first contacted the police for help. Individuals

who experience domestic violence appear to only resort to help beyond friends and family

in extreme circumstances because domestic violence is still considered largely a private

family matter.

Distrust of, and poor experience with, state authorities and public services also contributed

to not seeking help, a result that calls for further efforts to strengthen the services and

institutions aimed at supporting women and men affected by domestic violence in Ghana

– as well as coordination between them. Recommendations on moving forward are

provided at the end of the report, where detailed roadmaps are outlined on how best to

address the complex intersecting factors that underlie the incidence of domestic violence

among women and men and across different, socio-economic groups and locations across

Ghana.

20

1. Background to the study

Domestic violence – defined as abuse by one person against another in an intimate

relationship including marriage, cohabitation, dating or relations within the family – is one

of the most common forms of gender-based violence in the world (Abramsky et al., 2011;

Ellsberg et al., 2008; Garcia-Moreno et al., 2006, 2013; USAID, 2006). International studies

estimate that approximately 35 per cent of women across the world have experienced

physical and/or sexual violence at some point in their lives, largely in the form of domestic

violence (García-Moreno et al., 2013; Appendix A). 1 A recent systematic review has

reported that at least one in seven homicides and over one third of all female homicides

worldwide are perpetrated by an intimate partner (Stockl et al., 2013). Physical and sexual

violence are not the only types of domestic violence perpetrated against women:

emotionally abusive acts and controlling behaviour are experienced by up to 75 per cent

of women worldwide (García-Moreno et al., 2005). Domestic violence is also associated

with persistent forms of gender inequality and adverse health and economic outcomes

among victims, including poor physical and mental health, higher risks of the human

immunodeficiency virus (HIV) and other sexually transmitted diseases, restricted

livelihood options and choices, lower human capital and lower productivity (García-

Moreno et al., 2005, 2013; Moosa, 2012).

Civil society and governments around the world have acknowledged that violence against

women and girls (VAWG), including domestic violence, is a violation of basic human

rights and a global policy concern. In recognition of this international attention,

“eliminating all forms of violence against all women and girls in the public and private

spheres” is one of the targets against which Sustainable Development Goal number 5 –

aimed at achieving gender equality and empowering all women and girls – will be

measured. To further underscore the importance of tackling VAWG, the “elimination and

prevention of all forms of violence against women and girls” was chosen as the review

theme at the 60th Session of the United Nation’s (UN) Commission on the Status of Women

in March 2016.

Violence against women within private relations was drafted into international legislation

in the 1970s, as women’s movements in the United Kingdom and the United States of

America (USA) drew attention to the extent of violence committed by intimate partners.2

Domestic violence was first recognised at the international policy level in 1993 in the

1 This analysis is based on data from 155 studies in 81 countries. For further information on domestic violence

internationally, see also the Sexual Violence Research Initiative (http://www.svri.org/index.htm), and the

DFID-funded programme on What Works to Prevent Violence: A Global Programme to Prevent Violence

Against Women and Girls (http://www.whatworks.co.za/). Appendix A includes a comprehensive table

compiled by UN Women on the incidence of various forms of domestic violence across the world according

to the latest available information. 2 The first known use of the term ‘domestic violence’ in a political setting can be traced back to 1973, to an

address in the UK Parliament: in this address, domestic violence was explained to mean ‘spouse abuse,

violence in the home’: http://hansard.millbanksystems.com/commons/1973/jul/16/battered-wives. See also

Dobash and Dobash (1977), Freeman (1977) and Shanley (1993).

http://www.svri.org/index.htm

http://www.whatworks.co.za/

http://hansard.millbanksystems.com/commons/1973/jul/16/battered-wives

21

Declaration on the Elimination of Violence against Women (DEVAW), which framed

gender-based violence as a human rights violation.3 In these international declarations,

sexual and gender-based violence were defined primarily as violence committed by men

against women and girls. DEVAW, in particular, articulated three interlinked spheres in

which sexual and gender-based violence took place: in the family (including marital rape,

sexual abuse, female genital mutilation and dowry-related violence); in the community

(including rape, sexual harassment and sex trafficking); and by the State (all forms of

violence that are condoned or perpetrated by state actors).4 This approach was informed

by feminist theories that considered violence against women to be fundamentally different

from violence against men (Africa, 2010; Haraway, 1988; Yodanis, 2004). Emergent fields

of sexuality and masculinities studies have shown that domestic violence takes place

against women, against some men and also against children (Hunnicutt, 2009; see also

Institute of Development Studies et al., 2015).

This study provides a nationally representative, in-depth analysis of the incidence, levels

of acceptance, determinants and consequences of domestic violence, and the effectiveness

of institutional support offered to women and men who experience domestic violence in

Ghana. Section 1.1 below discusses in more detail current trends and research on domestic

violence in Ghana, while Section 1.2 describes the study.

1.1. Domestic violence in Ghana

The Republic of Ghana is located along the Gulf of Guinea in the sub-region of West Africa.

It has a total land area of 238,537 square kilometres and is bordered by Ivory Coast in the

west, Burkina Faso in the north, Togo in the east and the Atlantic Ocean in the south. Ghana

is divided into ten administrative regions: Western, Central, Greater Accra, Volta, Eastern,

Ashanti, Brong Ahafo, Northern, Upper East and Upper West.

3 Prior to this, human rights law did not cover gender-based violence because it excluded the ‘private sphere’

in which many forms of gender-based violence, such as domestic violence, took place. 4 1993 UN Declaration on the Elimination of Violence against Women.

22

Ghana’s population is around 24 million, according to the 2010 Census (Ghana Statistical

Service (GSS), 2013). Almost half of the total population live in the Ashanti, Eastern and

Greater Accra regions. Upper East is the least populated region, with 2 per cent of the total

population of Ghana. Over 50 per cent of the population live in urban areas (GSS, 2013).

The Ghanaian population is made up of several ethnic groups. The Akans are the largest

group (48 per cent), followed by the Mole-Dagbani (17 per cent), Ewe (14 per cent), Ga-

Dangme (7 per cent) and others (GSS, 2013). The sex ratio, according to the 2010 Census, is

95.2 males per 100 females.

According to the Demographic and Health Survey (DHS) conducted in Ghana in 2008, 38.7

per cent of ever-married women between the ages of 15 and 49 years reported having

experienced physical, psychological or sexual violence by a husband or partner at some

point in their lives. Over a quarter (27.6 per cent) of Ghanaian males reported having

experienced physical or psychological violence by their wife or partner (GSS et al., 2009).

Considerable effort has been made in Ghana over the last three decades to reduce the

incidence of domestic violence. Some of the first studies on domestic violence in Africa

took place in the 1990s in Ghana, as well as in Tanzania, Uganda and South Africa

(Hodgson, 2002; Ofei-Aboagye, 1994; Watts, Osam and Win, 1995). These studies were

motivated by the actions of activist groups, which played an important role in the

formulation and passing of domestic violence laws in Ghana (Bowman, 2002; Kimuna and

Djamba, 2008; Schneider, 2008).

23

Ghana’s first legislation efforts against domestic violence reflected the global momentum

in pushing for women’s rights to be recognised as human rights (Cook, 2011),5 prompted

by years of advocacy from key civil society organisations (CSOs), women’s rights

organisations and international bodies. These actions resulted in the Government of Ghana

enacting a number of national laws to protect women’s rights and outlaw violence against

women and girls. These included a provision in the 1992 Constitution that prohibited

discrimination based on sex; the 1998 Criminal Code Amendment Act; 6 and legal

amendments criminalising certain harmful traditional practices, such as widowhood rites

(1984), female genital mutilation (FGM) (1995) and child abuse (1998). In February 2007,

Ghana’s Parliament voted to enact the Domestic Violence Act (Act 732). Appendix B

provides a detailed historical timeline of international and Ghanaian policies related to

domestic violence.

The enactment of the 2007 Domestic Violence Act was the culmination of a long process of

advocacy and civil action, which started with the publication of the Gender Studies and

Human Rights Documentation Centre’s (Gender Centre’s) research on violence against

women and children, by Coker-Appiah and Cusack (1999), and continued under the

leadership of the Gender Centre, the Women’s Initiative for Self-Empowerment (WISE)

and the Ark Foundation.7 Responses to domestic violence were largely initiated by non-

governmental organisations (NGOs), CSOs and networks and coalitions in Ghana that

worked in the areas of human rights and women’s rights (Amoakohene, 2004; NETRIGHT,

2012).8 Together, these institutions have been engaged in activities to change perceptions,

attitudes and harmful social norms among local populations and within public institutions.

The efforts to put the Domestic Violence Act in place in the late 1990s were accompanied

by much publicity in the media. Physical spousal abuse became a regular theme in media

reports, which resulted in the establishment of the Women and Juvenile Unit (WAJU) of

the Ghana Police in 1998, a specialised unit that handled crimes against women and

children (Amoakohene, 2004). During early 2002, an increased number of stories on a series

of murders of women by their partners were publicised, fuelling further the debate on

violence against women in Ghana and enabling CSOs to mobilise public action. The role

5 These include the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW),

the African Charter on Human and Peoples’ Rights (ACHPR) and the Maputo Protocol on the Rights of

Women in Africa. 6 This Act amended the Criminal Code of 1960 by increasing the age of criminal and sexual responsibility,

including a specific offence of indecent assault and revising provisions regarding sexual offences. It also

abolished customary or ritual servitude. 7 WISE provides counselling and other support services to survivors of domestic violence and has played a

central role in increasing public awareness about domestic violence. The Ark Foundation was instrumental

in forming various gender coalitions, such as the Sister’s Keepers, the Gender Violence Survivors’ Support

Network (GVSSN), the Network for Women’s Rights (NETRIGHT) and the Violence against Women and

Children’s Practitioners Network, among others. 8 These included, among others, the Ghana branch of the International Federation of Women Lawyers (FIDA-

Ghana), which worked together with other NGOs to officially abolish trokosi (a traditional religious practice

whereby human beings, usually young virgin girls, are used as payment to the Gods) and to lobby the

government to pass the Domestic Violence Bill, among other initiatives.

24

of the media at that time was criticised for their sensational framing of events and ill-

considered language that left the impression that domestic violence was something

unusual, based on individual malfunctioning rather than being situated within wider

cultural and institutional norms and practices. But Gadzepko (2009: 266) also describes the

mass media as “a key channel of information and debate on domestic violence and

consequently [...] influential in constructing public understanding of violence against

women and children”.

Efforts by civil society actors and NGOs resulted in successful mobilisation, advocacy,

sensitisation and education about domestic violence, and the training of state officials on

domestic violence and gender issues. Their efforts achieved momentum with the coming

together of stakeholder CSOs into the National Coalition on Domestic Violence Legislation,

established in 2003, which “at various times worked closely with, and at other times

independently of, or even in conflict with, the State” (Adomako Ampofo, 2008: 1).

Finally, in 2007, Ghana’s Parliament voted to enact the Domestic Violence Act (Act 732).

The actual Bill had been before Parliament since 2003 and was the “subject of heated

debate” (Adomako Ampofo, 2008). According to Takyiwaa Manuh, the “process leading

to the passage of the law involved not only the introduction of new legislation, but also

confronting a social system that tolerates various forms of violence against women and

girls, especially in the context of gender relations and in the domestic sphere” (Manuh,

2007: 1).

Ghana’s domestic violence legislation takes a broader and, arguably, culturally sensitive

approach to access to justice, when compared with other countries. First, the 2007 Domestic

Violence Act allows for mediation by alternative dispute resolution methods.9 Second, the

definition of domestic violence used in Act 732 does not include reference to a specific sex.

Third, the Act acknowledges that perpetrators and survivors do not have to be married or

related by blood ties, and applies to live-in household staff too. Fourth, the definition of

domestic violence includes various forms of economic abuse, in addition to more

conventional definitions of sexual and physical violence. The Act also provides a working

definition of domestic violence, and outlines a comprehensive legal framework for the

prevention of and protection against domestic violence. Notably, the law criminalises

various acts of physical and sexual violence, economic and psychological abuse, and

intimidation and harassment in domestic relations. These broad definitions of domestic

violence have implications for research on the incidence, nature, causes and consequences

of domestic violence in Ghana, which will be discussed in more detail in the next section.

9 Act 732 allows for the Court, with the consent of the complainant, to refer the case for settlement by an

alternative dispute resolution method if the criminal trial of a domestic violence case is not aggravated; if the

complainant requests to have the matter settled out of court; or if the court determines that the case “can be

amicably settled”. Local mediation, however, entails its own challenges, which will be discussed in Section

4.5.4.

25

1.2. The study

The Ghana Domestic Violence Act (Act 732) was followed by the formulation of the

National Policy and Plan of Action (NPPOA) developed by the former Ministry of Women

and Children’s Affairs in 2008. The NPPOA lays out the specific roles of key stakeholders

that will effectively implement the Domestic Violence Act, under the direction of the

Domestic Violence Secretariat, part of the Ministry of Gender, Children and Social

Protection (MoGCSP). This study forms part of this strategy. The overarching aim of the

study is to provide an in-depth understanding of the attitudes to and incidence,

determinants and consequences of domestic violence in Ghana. The study addresses the

following questions:

1. What types and levels of violence are experienced by women, men, girls and boys

in Ghana, and who perpetrates them? What proportion of this violence is domestic

violence?

2. What are society, community and people’s attitudes towards domestic violence, and

how do attitudes and norms influence the acceptance of different forms of domestic

violence in Ghana?

3. What are the main determinants of domestic violence against women, men, girls

and boys at the individual, family, community and wider societal levels?

4. What are the consequences of domestic violence in Ghana for people’s physical and

mental health and their daily lives?

5. What institutional structures support victims of domestic violence, and what

proportion of victims are able to effectively access available services?

6. How can policies and interventions be (re)designed to better support victims of

domestic violence and prevent domestic violence in Ghana?

The study adopted a mixed-methods approach, which involved the collection of primary

quantitative data through a representative household-level survey of 4,995 individuals –

the Ghana Family Life and Health Survey (GFLHS) 2015 – and the collection of qualitative

in-depth data on key mechanisms that may explain trends and the distribution of domestic

violence outcomes in Ghana.

Qualitative and quantitative methods were fully integrated to enable complementarity

between the methods. The quantitative data enabled the production of statistics that are

regionally and nationally representative at the household level, as well as an analysis of

determinants and consequences of domestic violence. The qualitative data explored the

perceptions, attitudes and beliefs of women and men within their everyday lives, how they

are related to domestic violence experiences as survivors and perpetrators, and how

individual experiences of domestic violence are integrated within family, community and

wider social and political structures and relations.

The next sections present the detailed findings of this study. Section 2 discusses the main

definitions of domestic violence adopted in the study, and outlines the conceptual

26

framework of the report. Section 3 explains the qualitative and quantitative methodologies

used in the study to collect nationally representative data on domestic violence in Ghana.

Section 4 discusses in detail the main findings of the study. Section 5 summarises the key

findings and discusses them in relation to previous studies conducted internationally and

in Ghana. Section 6 concludes with proposals for future policy action to reduce and prevent


27

2. Conceptual framework

Five main sets of factors have historically been used to describe the determinants of

domestic violence: individual psychological factors, sociological factors (such as family

dysfunction), male aggression, poverty and the prevalence of cultures of violence

(Bowman and Schneider, 1998; Schneider, 2000).10 International studies, largely conducted

in the USA, have emphasised the role of individual psychology, interpersonal relationships

within families and male aggression as the main explanations for the prevalence of

domestic violence (Gosselin, 2000; Heath, 2014; Perilla, Bakeman and Norris, 1994; Pollak,

2004; Schneider, 2000; Walker, 1999). These individual- and family-centric approaches to

domestic violence have been criticised for their lack of focus on factors such as poverty and

culture – arguably more relevant in developing countries. They also tend to prioritise

response mechanisms, such as psycho- and family therapy (Pollak, 2004), crime control

and retroactive responses to cater to survivors’ needs (Kelly, 2003), which may not be

useful in settings where medical and legal recourse may be prohibitively expensive and

difficult to access (Mann, 2000; Murphy, 2002). Approaches that rely on expensive therapy

to facilitate transformation may not, in addition, account for the deeper structural

inequalities that have been shown to contribute to domestic violence (Maguigan, 2002;

Schneider, 2008).

As a response, feminist theories have attempted to analyse domestic violence as an

outcome of broader contexts characterised by unequal gender relations (Bowman, 2002).

This approach has been prolific in studies on domestic violence in Africa (Banda, 1995; also

see Adjetey, 1994; Bell and Mattis, 2000; Bograd, 1999; Butegwa, 1994; Koenig et al., 2003;

Perilla et al., 1994; Sokoloff and Dupont, 2005; Tamale, 1993; Turshen and Turshen, 2000),

including one of the first domestic violence studies in Ghana by Rosemary Ofei-Aboagye

(Ofei-Aboagye, 1994a; also see Akpalu et al., 2000). These studies have found that domestic

violence against women is largely explained by women’s subordination to, and economic

dependence on, men.

Feminist theories are closely linked to cultural approaches to domestic violence. Among

these, several studies have found that some aspects of domestic violence – for instance,

wife-battering – are considered ‘normal’ within particular cultures. As an example, a study

in Nigeria revealed that police officers tended to remind wives who reported physical

violence that their ‘Yoruba culture’ permits men to beat women (Atinmo, 1997; Atinmo,

2001). Other less direct examples of ‘culture’ as an explanation for domestic violence

include the widespread practice of bride price, and gender norms that limit women’s

10 The study was informed by a comprehensive literature review and secondary data analysis. Electronic key

word searches were initially carried out in the English language in Science Direct, Scopus and Google Scholar.

Subsequently, searches for regional and international policy documents were conducted using the United

Nations and the African Commission websites. Thereafter, an extended search was done with academic

publications by following ‘related’ or ‘cited by’ articles offered on publishing journals’ websites and Google

Scholar. Studies that could not be obtained online or as hard copies in the United Kingdom were collected

during visits to Accra.

28

access to work outside their home (Abane, 2003; Armstrong, 1998). For instance, research

in Zimbabwe has found that asking for money from male partners led to physical violence

against women, as the request was interpreted as a challenge to men’s position of power

(Armstrong, 1998). Similarly, in Ghana, Amoakohene (2004: 2378) has analysed how

women are expected “to be ‘submissive to [their] husbands’, ‘be respectful’, ‘be patient’,

and ‘be dutiful and serviceable’, as tradition demands […]”.

A related approach has argued that domestic violence is explained (at least partially) by

the shift from traditional to modern, urbanised societies, a process that may result in some

men feeling threatened and more likely to find ways of asserting their power through

violence (Bowman, 2002). Abane (2003) has found that socio-economic change in Ghana

has been accompanied by increasing economic strain within households (see also

Armstrong, 1998). As a result, some men faced increased difficulties in supporting multiple

wives and households, while women took on waged employment to subsidise the family’s

income. However, women were expected to do both household and paid work. If they

failed to meet their partner’s expectations, they could be beaten. Amoakohene (2004), too,

has found that a change from traditional gender roles, where women worked at home, to

new roles, where women took on waged employment, precipitated increases in domestic

violence. International studies have revealed similar results (Justino et al., 2012).

These approaches have resulted in important insights into the causes and consequences of

domestic violence. However, most emphasise a unidimensional perspective, either

focusing on the individual, on immediate social relations or on wider social and political

processes. Structuralist theories, in contrast, have attempted to identify how drivers and

responses to domestic violence may be shaped simultaneously by social, political,

economic and cultural structures. Historically, structural approaches were championed by

some feminists in the 1970s and 1980s who, in line with CEDAW and DEVAW, argued that

domestic violence was not a private matter between individuals in the hidden spaces of

their homes. Domestic violence should, rather, be understood as a public issue shaped by

state and community actors and institutions that privilege certain gender and sexual norms

over others (Bowman and Schneider, 1998).

Within this theoretical perspective, the social ecological model (Bronfenbrenner, 1979;

Carlson, 1984; Heise, 1998; Heise and Garcia-Moreno, 2002) offered a conceptual approach

that integrated many of the above theories into a structural model that explored domestic

violence across individual, societal, economic and political spheres. The model outlined

four key levels of interaction across public and private spaces: the individual (personal

history), the family (ecosystem), the community (microsystem) and the social and political

context (macrosystem).

Subsequent critiques led to the refinement of the original model. The first critique related

to the model presenting the four layers as impermeable and unrelated to one other.

However, studies in Brazil (Minayo, 1994), South Africa (Jewkes et al., 2002) and the USA

(Perilla et al., 1994) highlighted the extent to which the experiences and personal histories

29

of the individual were strongly entwined with the ‘ecosystem’, their workplaces and

neighbourhoods. The second critique originated from post-structuralist and post-colonial

theorists who emphasised the importance of seeing individuals and structures in a circular

dynamic, where people have constrained agency within structures that they both

reproduce and challenge (Bourdieu, 2001). These theories resonated with post-colonial

critiques, particularly by African feminists (Jones, 2011; Coughtry, 2011; Ewelukwa, 2002;

Tsikata, 2014). These argued that both individualist and social ecological approaches were

ahistorical and Western-centric. Furthermore, they failed to account for the role of violently

unequal economic and social systems – such as colonialism, apartheid and globalisation –

in entrenching intersecting forms of violence, including domestic violence, which were

linked to class, race, sex, sexuality, age, location and so forth (Africa, 2010; Cornwall et al.,

2007; Jones, 2011; McFadden, 2000).

These critiques point to the importance of a conceptual framework that considers through

an explicit political economy lens the complex interplay between the individual, the family,

the community and the wider political and social system in Ghana. This study has,

therefore, adopted a ‘critical social ecological’ approach as an amalgamation of the original

social ecological theory adapted by Carlson (1984), Heise (1998) and others to domestic

violence, and the extensions proposed by Jewkes (2002). At a general level, this integrated

approach requires an understanding of individual experiences of domestic violence within

family relations, the immediate political and economic environment, local institutions

(such as the police and health care providers) and community attitudes towards violence,

and the current and historical national policy context and political and economic

environment. These four spheres and their interactions are outlined below.

The individual. Personal circumstances help to understand the attitudes, incidence,

determinants and consequences of domestic violence. As discussed above, domestic

violence often goes unrecognised, unreported and denied due to the reluctance of

individuals to report ‘private’ matters. One important example is the extent of sexual abuse

within marital relationships (Adomako Ampofo, 2008). Other examples include human

trafficking, coerced sex work, child marriage, widow inheritance and female genital

mutilation (FGM). The international literature on domestic violence has shown that

personal histories contribute to experiencing or enacting violence. Examples include

witnessing violence as a child, having experienced abuse as a child or having had an absent

father (García-Moreno et al., 2005; WHO, 2000; UNICEF, 2002). Based on these findings,

this study was designed to explore people’s experiences with domestic violence over time,

through survey questions, focus group discussions and in-depth interviews. These

methods have been able to identify people’s personal experiences of domestic violence,

their attitudes towards domestic violence, and their family and social relationships that

may account for these experiences and attitudes – from the perspectives of both survivors

and perpetrators.

The family. The family is the immediate environment in which people who experience or

exert violence live. In this sphere, factors that contribute to domestic violence include male

30

dominance and male control over the family’s wealth. These factors may also work

conversely: if men do not hold wealth and women do, women may be subject to violence

by her partner, or even by her partner’s family (Armstrong, 1998; Geisler, 1995; Njovana

and Watts, 1996). The 2008 Ghana DHS study found that the incidence of domestic violence

was associated with a complex variety of causes within the nuclear and extended family

network. The results revealed that 36.6 per cent of women aged 15–49 years experienced

physical violence, mostly perpetrated within the family by current or previous husbands

or partners, as well as by other male and female family members or acquaintances (GSS et

al., 2009). The current study, therefore, incorporated methodologies to collect data on and

analyse domestic violence from this family perspective. To this purpose, the survey

questionnaire included several questions on the determinants of violence within the

domestic sphere, and about social norms and expectations around gender roles within

immediate and wider family circles. The qualitative instruments were designed to examine

personal experiences within families, by allowing interviewees to tell personal stories

about relevant events, and explore different elements in the story to understand what

happened and why.

The community. This level of analysis takes into account the institutions and social

structures that surround the family, such as the neighbourhood, social networks and the

workplace. Human perceptions and attitudes do not emerge in a vacuum, and

communities perform important functions in the lives of individuals and their families.

Notably, the attitudes of community members, traditional and religious leaders,

neighbours, work groups and other local social groups and networks are likely to shape

the attitudes of individuals and their families, including the acceptance or tolerance of

domestic violence. In Ghana, key community figures include community leaders and local

representatives of state agencies, such as police officers and health care workers. This

current study has paid careful attention to the analysis of how attitudes and norms towards

domestic violence are shaped by wider community-level social, economic and political

dynamics. The surveys for this study were designed to capture different types of violence

experienced within the environment surrounding an individual (notably, the workplace

and the community). The survey and qualitative instruments also included questions on

attitudes and norms about the acceptability of domestic violence, about appropriate

reactions and potential punishments, and about beliefs that may affect people’s ability to

take cases of domestic violence outside the private sphere and to involve state actors, such

as the police.

The wider social and political context. Social and political contexts shape general views

and attitudes shared across societies. The original social ecological model applied to

domestic violence paid limited attention to these wider political or economic structures.

Later work by Jewkes et al. (2002) revealed how political and social institutions engage,

reproduce and challenge gender inequalities and domestic violence, and how economic

inequalities affect people’s experiences of domestic violence. Key political and social

institutions include the police, law courts, legal services, and health and social support

organisations. This is because domestic violence may go unchecked if not accompanied by

31

integrated prevention and oversight through the State’s law enforcement bodies. Domestic

violence may also be unreported if it is socially accepted across the wider society. One

objective of the current study was to uncover how institutions set up to respond to and

prevent domestic violence in Ghana may have shaped the incidence, nature, determinants

and consequences of domestic violence across the individual, their families, their

communities and the wider society. Quantitative and qualitative instruments have

measured differences in knowledge, availability, accessibility, use and satisfaction with

different services related to domestic violence. Information was also collected about views

and attitudes across the population that may explain relations between institutional

engagement, structural inequalities, gender inequality and domestic violence. The

qualitative instruments, in particular, collected data on whether people were aware of

available services, the barriers they might have encountered when accessing them, and

suggestions to support individuals and communities to hold local and national

government accountable to the provision of services listed in the 2007 Domestic Violence

Act.

The conceptual framework highlights the importance of understanding domestic violence

both within and across the different spheres that shape the incidence, distribution and

nature of domestic violence: the individual, their immediate social relations, the

community and the wider social and political system. The study used this conceptual

framework to collect new empirical evidence across Ghana on: (i) the incidence and types

of domestic violence; (ii) individual-, household- and community-level attitudes towards

domestic violence; (iii) determinants of domestic violence; (iv) consequences of domestic

violence; and (v) the use of and access to victims’ support services and institutions.

Subsection 2.1 discusses the main definitions and concepts used in the study. Subsections

2.2, 2.3, 2.4 and 2.5 describe, respectively, the current study’s approach to the collection

and analysis of information on attitudes to and determinants and consequences of

domestic violence, and how people access and use relevant services and institutions across

Ghana.

2.1. Types and perpetration of domestic violence

The most prominent international studies on domestic violence are of a quantitative

nature, based on national demographic data and health surveys (e.g. UNICEF, 2002; WHO,

2000). These typically measure physical, sexual and psychological abuse by using a set of

questions about particular acts of violence and asking respondents to state whether they

have ever experienced them, and if so, whether this happened in the 12 months prior to

the survey. These studies also usually investigate attitudes towards physical and sexual

abuse by men and women.

Studies prior to 2008 in Ghana were of a more exploratory nature but included very useful

information about context-specific and normative understandings of what constituted

domestic violence. The first study on the nature and extent of domestic violence in Ghana,

by Rosemary Ofei-Aboagye (1994), focused on women’s understandings and responses to

32

domestic violence. The study noted, in particular, the importance of terminology: only five

out of 50 women said it was accepted in their ethnic group that husbands beat their wives;

but many more did so when the term ‘disciplining’ was used. A second study by the

Gender Centre was designed to better understand the prevalence, type, contexts,

responses, barriers to responses and recommendations about domestic violence at the

national level (Coker-Appiah and Cusack, 1999). Similar to Ofei-Aboagye’s 1994 study, the

researchers asked women to define what they understood as violence. The study revealed

that most women understood that violence was not limited to physical abuse but also

included acts of psychological mistreatment such as threats, humiliations, being prevented

from seeing family and friends and sexual harassment. A later study by Amoakohene

(2004) analysed domestic violence from a more multidimensional perspective, by

exploring women’s perceptions about their rights, responsibilities, duties and the abuses

they suffer, as well as the coping strategies they developed. Domestic violence was seen to

include both physical and psychological abuse. Sexual violence was not recognised as an

offence by either the State or the community (see also Adu-Gyamfi, 2014).

These findings, alongside the definition of domestic violence proposed in the 2007

Domestic Violence Act, have informed the typology of domestic violence used in this

study. Notably, the 2007 Domestic Violence Act defines domestic violence as: “engaging in

the following within the context of a previous or existing domestic relationship:

(a) an act under the Criminal Code 1960 (Act 29) which constitutes a threat or harm to a

person under that Act;

(b) specific acts, threats to commit, or acts likely to result in

(i) physical abuse, namely physical assault or use of physical force against another person

including the forcible confinement or detention of another person and the deprivation of

another person of access to adequate food, water, clothing, shelter, rest, or subjecting another

person to torture or other cruel, inhuman or degrading treatment or punishment;

(ii) sexual abuse, namely the forceful engagement of another person in a sexual contact

which includes sexual conduct that abuses, humiliates or degrades the other person or

otherwise violates another person's sexual integrity or a sexual contact by a person aware of

being infected with human immunodeficiency virus (HIV) or any other sexually transmitted

disease with another person without that other person being given prior information of the

infection;

(iii) economic abuse, namely the deprivation or threatened deprivation of economic or

financial resources which a person is entitled to by law, the disposition or threatened

disposition of moveable or immovable property in which another person has a material

interest and hiding or hindering the use of property or damaging or destroying property in

which another person has a material interest; and

(iv) emotional, verbal or psychological abuse namely any conduct that makes another

person feel constantly unhappy, miserable, humiliated, ridiculed, afraid, jittery or depressed

or to feel inadequate or worthless;

(c) harassment including sexual harassment and intimidation by inducing fear in another

person; and

(d) behaviour or conduct that in any way

33

(i) harms or may harm another person,

(ii) endangers the safety, health or well-being of another person,

(iii) undermines another person's privacy, integrity or security, or

(iv) detracts or is likely to detract from another person's dignity and worth as a human

being.

Following this definition, the study has collected detailed information on all categories of

physical, sexual, economic and psychological abuse. In line with prior studies conducted

in Ghana and information provided to the research team in initial scoping interviews with

key stakeholders, the study also collected information on acts of controlling behaviour

(social violence). Detailed definitions of the types of violence are provided in Section 4.

Conforming to international definitions (e.g. UNICEF, 2002; WHO, 2000), the Domestic

Violence Act, and prior studies conducted in Ghana, domestic violence is defined in this

study as violence committed against a person by an intimate partner, including co-habiting

partners, and by other family members, irrespective of whether this violence occurs within

or beyond the confines of the home.

Ghana’s Domestic Violence Act provides a broad definition of what constitutes a ‘domestic

relation’, which has been adopted in this study. ‘Domestic relation’ means “a family

relationship, a relationship akin to a family relationship or a relationship in a domestic situation

that exists or has existed between a complainant and a respondent and includes a relationship where

the complainant (a) is or has been married to the respondent; (b) lives with the respondent in a

relationship in the nature of a marriage even if they are not or were not married to each other or

could not or cannot be married to each other; (c) is engaged to the respondent, courting the

respondent or is in an actual or perceived romantic, intimate, or cordial relationship not necessarily

including a sexual relationship with the respondent; (d) and respondent are parents of a child, are

expecting a child together or are foster parents of a child; (f) and respondent are family members

related by blood ties, affinity or adoption, or would be so related if they were married either

customarily or under an enactment or were able to be married or if they were living together as

spouses although they are not married; (g) and respondent share or shared the same residence or are

cotenants; (h) is a parent, an elderly blood relation or is an elderly person who is by law a relation

of the respondent; (i) is a house help in the household of the respondent; or (j) is in a relationship

determined by the court to be a domestic relationship”.

To distinguish whether acts of violence are domestic or non-domestic in nature,

respondents who reported having experienced any one of the acts of violence above during

the 12 months prior to the survey were asked whether the perpetrator was part, or not, of

the household. If the perpetrator and survivor shared the same roof and meals (the

definition of a household adopted in this study, following the standard definition adopted

by GSS), the respondent was asked to identify the perpetrator. In cases where the

perpetrator was identified as living outside the household, respondents were asked about

their relationship to the perpetrator. Answer categories included (former) partners,

extended family members living within or outside the same community, non-related

34

community members, teachers, colleagues, friends and other relations. This choice of

questions allowed the study to identify which violence was domestic in nature, as well as

to frame experiences of violence within the wider community and societal contexts.

One innovative aspect of the 2007 Domestic Violence Act is that its definition of domestic

violence does not refer exclusively to women. Although domestic violence is generally

disproportionately directed at women, some studies – almost exclusively based on data

from the developed world – have argued that this is not always the case: women can be

equally violent, and children also ought to be considered potential victims of domestic

violence (Hunnicutt, 2009). The DHS 2008 study conducted in Ghana was one of the few

studies to date to include a domestic violence module that elicited men’s experiences. Over

a quarter (27.6 per cent) of ever-married Ghanaian men aged 15–49 years in the 2008 DHS

reported having experienced physical or psychological violence by their wife or partner at

some point during their lives (GSS et al., 2009). Using the same data, Kishor and Bradley

(2012) found that 19 per cent of married women and 10 per cent of married men had

experienced physical, psychological or sexual spousal violence, confirming that it is not

only women who are victims of intimate partner violence.11 In addition, the Domestic

Violence Victims Support Unit (DOVVSU) reported to the research team during interviews

in Ghana that it had processed several cases reported by men and boys. At the international

level, similar evidence is reported in IDS et al. (2015).12 This study has attempted to uncover

the experiences of women and girls, as well as those of men and boys. It is also important

to note that different types of domestic violence and reporting behaviour of domestic

violence vary across different age groups, regions and socio-economic characteristics

(Coker-Appiah and Cusack, 1999; García-Moreno et al., 2005; GSS et al., 2009). The study

will address these differences in detail in Section 4.

2.2. Attitudes towards domestic violence

Similar to other social attitudes, attitudes towards domestic violence depend on a range of

factors. These include individuals’ beliefs about gender norms and norms that regulate the

use of violence; how such beliefs are internalised; social norms that regulate the

distribution of power and resources within the household and immediate family; and

social norms that shape gender roles and behaviours at the community and wider societal

levels (Justino, Leone and Salardi, 2015). Multi-country studies and international

comparisons have shown that women’s attitudes towards violence vary substantially

across settings. Furthermore, women living in countries where domestic violence is more

prevalent tend to find domestic violence more acceptable than women in countries where

prevalence is lower (García-Moreno et al., 2005; Appendix A). The acceptance of wife-

11 However, the authors show that women are consistently and significantly more likely than men to

experience all forms (physical, sexual and emotional) of such violence, and that intimate partner violence is

more commonly experienced by women. It is also more severe and more likely to result in injuries than

intimate partner violence experienced by men. 12 See also the research programme on Men, Boys and Gender Equality at the Institute of Development

Studies: http://menandboys.ids.ac.uk/.

http://menandboys.ids.ac.uk/

35

beating also tends to be higher among women who have experienced abuse than among

those who have not (García-Moreno et al., 2005). Taken together, these findings suggest

that women seem to learn to accept violence when they are victims themselves – a result

that will be explored further in the study.

Ofei-Aboagye’s (1994) study asked respondents about the level of beating that was

perceived to be acceptable. The findings showed that one to three slaps were considered

acceptable. However, the study also found that there was a threshold beyond which

beating ‘transcended the norm’. This happened when the victim was permanently hurt,

scarred or deformed, or was beaten in public. A later study by Takyi and Mann (2006)

found that 40.5 per cent of never-married and 33 per cent of ever-married men perceived

wife-beating to be justified. The most justifiable reason given was when they perceived

that women neglected their children. Using data from the 2008 DHS, Mann and Takyi

(2009) found that women were more likely than men to believe that spousal violence

against them was justified: 64.2 per cent of men stated that abuse would never be justified,

but only 44.6 per cent of women agreed (see also Kishor and Bradley, 2012; Tenkorang et

al., 2013). These findings raise important questions about the prevalence of social norms

and beliefs that may sustain domestic violence – an issue that will be explored in detail in

Section 4.

It appears that wife-beating may be becoming less accepted. Ghana DHS statistics for 2003

and 2008 show that, across almost all regions, the percentage of respondents condoning

wife-beating decreased. The reductions were up to 35.7 percentage points among men, and

up to 45.9 percentage points among women, with the largest changes taking place in the

Upper East and Upper West regions. These were also regions with a traditionally high

prevalence of domestic violence, and where awareness-raising campaigns have

concentrated their sensitisation efforts. This study will explore how these trends may have

continued until 2015 and what social norms and changes in social norms may explain them.

2.3. Determinants of domestic violence

The existing literature on domestic violence has found evidence for strong correlations

between the incidence of domestic violence and a wide range of individual socio-economic

factors, such as age, education and marital status (Aizer, 2011; García-Moreno et al., 2005;

Jewkes et al., 2002; La Mattina, 2012). In line with the conceptual framework outlined

above, evidence shows that these individual determinants of domestic violence interact

with factors that play out at the level of families (notably, differences in intra-household

gender dynamics), as well as wider community and societal levels – in particular, the

prevalence of patriarchal norms, perceptions of masculinity and exposure to other forms

of violence. These determinants of domestic violence, and how the study was designed to

capture them, are discussed below.

Age and intergenerational effects. Several studies have shown a close association between

the incidence of domestic violence and the age of victims and perpetrators, with, in general,

36

incidence being lower among older women over their lifetime and in the recall period of

the surveys (Aizer, 2011). For example, in urban Peru, 41 per cent of women aged 15–19

years old reported having experienced intimate partner violence, in comparison to 8 per

cent among women aged 45–49 years old (García-Moreno et al., 2005). The research team

conducted a background analysis of the Ghana DHS 2008 for this study. In contrast to

international evidence, the results did not show any particular age patterns among victims

of domestic violence. The data indicated, however, that the incidence of domestic violence

in Ghana was higher among women and men whose father beat their mother. Several other

studies have found exposure to domestic violence during childhood to be positively

associated with the probability of being a victim of domestic violence as an adult (Clark et

al., 2010; Gallegos and Gutierrez, 2011; Bobonis et al., 2013). These findings are consistent

with theories of intergenerational transmission of domestic violence (Pollak, 2004). This

study will investigate further the distribution of the incidence of domestic violence across

age and generations in Section 4.

Marital status. Qualitative research has shown that, in many parts of the world, married

men are assumed to have unconditional sexual access to their wives (Ellsberg et al., 2001).

In the context of Ghana, Adu-Gyamfi (2014) has reported how attitudes, beliefs and

practices have led to violent actions against women being dismissed due to widespread

beliefs that women give up their rights to their husbands at the point of marriage,

particularly when bride prices are paid. In the international literature, marital status is also

an important factor associated with domestic violence. Women who are separated or

divorced tend to report higher incidence of domestic violence than those who are currently

married, with violence persisting after the divorce or separation (García-Moreno et al.,

2005). The timing of marriage also appears to be an important determinant of domestic

violence, with the incidence of domestic violence lower among victims who marry later

(La Mattina, 2012). The background analysis of the 2008 DHS data conducted for this study

showed that the incidence of domestic violence in Ghana was lower among currently

married individuals (women and men) than among those who were formerly married and

separated or divorced afterwards. The role of marital status on domestic violence in Ghana

in 2015 will be explored further in Section 4.

Socio-economic status and education levels. Though domestic violence affects everyone,

existing literature indicates that the incidence of domestic violence is usually associated

with low income and wealth levels (Bobonis et al., 2013). For instance, Aizer (2011) showed

that poor women were at a higher risk of abuse in the USA. In contrast, Amoakohene (2004)

reported domestic violence in Ghana to be widespread among all socio-economic classes,

with 70 per cent of educated and employed respondents in her sample reporting some

form of abuse (predominantly physical). The study by Adinkrah (2014) showed that

homicide-suicides occurred across the whole socio-economic spectrum but were more

prevalent among individuals with lower socio-economic background. Education levels are

also typically associated with incidence of domestic violence. As argued in Jewkes et al.

(2002), “education confers on individuals social empowerment, self-confidence, and the

ability to use information and resources to one’s advantage”. Aizer (2011) found that less

37

educated women were at a higher risk of abuse in the USA, while Panda and Agarwal

(2005) showed that women who owned land in Kerala (India) were less likely to experience

forms of domestic violence. García-Moreno et al. (2005) reported that in many of the

countries they analysed the prevalence of domestic violence decreased with rising

education levels. However, in some of the countries (Brazil, Namibia, Peru, Thailand and

Tanzania) the prevalence of domestic violence was lower only among women with

education levels above secondary school. This may imply that, at higher levels of

education, women acquire a greater capacity to negotiate autonomy and control of

resources within their marriage or families. Consistent with these findings, Tenkorang et

al. (2013) have shown that women with higher levels of education in Ghana in 2008 were

less likely than women with lower levels of education to experience physical violence.

Yeboah and Batse (2009) found some level of schooling (primary education) to be

correlated with higher risks of psychological violence. Education has also been found to

explain attitudes towards domestic violence: Takyi and Mann (2006) reported that younger

men and men with little education or social status are more likely than other socio-

economic groups to believe that wife-beating is justified.

According to the 2008 Ghana DHS, the incidence of domestic violence in Ghana was lower

among individuals with no education than among those with primary or higher levels of

education. La Mattina (2012) found a similar result in Rwanda. These results could suggest

that individuals who are more educated may also be more likely to be aware of existing

laws against domestic violence and to report being victims of domestic violence. The 2008

data for Ghana showed that the percentage of individuals who condoned wife-beating was

higher among those with no education than among those with some level of education.

The role of socio-economic status, and education in particular, will be further explored in

Section 4.

Intra-household gender dynamics. Several studies have attributed changes in rates of

domestic violence to changes in gender dynamics within the household. A corollary of the

feminist theories discussed in the previous section is that women who have more

autonomy and earn independent incomes may be less likely than other women to

experience domestic violence (see Goode, 1971; Dyson and Moore, 1983). For instance,

Aizer (2010) showed that increases in women’s relative wages and in women’s labour force

participation resulted in a lower incidence of domestic violence in the USA. In contrast,

others have noted that newly economically empowered women may threaten men’s

identity, resulting in increases in domestic violence. In these cases, husbands may resort to

violence to reassert their sense of power (Heath, 2014; Menon and Johnson, 2007; Tanchen,

Witte and Long, 1991). Consistent with these predictions, studies have found that increases

in women’s labour force opportunities and their access to income and other resources have

at times been associated with increases in domestic violence. This is particularly true if

women’s initial level of bargaining power before entering the labour market was low

(Bobonis et al., 2013; Heath, 2014; Hidrobo and Fernald, 2013; Justino et al. 2012).

38

In the case of Ghana, Mann and Takyi (2009) have shown that the likelihood of an

individual tolerating domestic violence was 50 per cent higher among men who

contributed less than half of the household budget (when compared to men who

contributed to the totality of the household budget). Interestingly, when men and women

were equals in terms of decision-making, men were more likely than women to believe

that abusing women was not justifiable. The background analysis conducted in this study

using the 2008 DHS data also showed that the incidence of domestic violence was lower

when men and women made decisions jointly. The incidence of domestic violence was

higher among women who made decisions about household expenses alone. These

patterns will be explored further for the 2015 data in Section 4.

Patriarchal norms. Several theories have proposed patriarchy as an explanation for the

prevalence of domestic violence, whereby violence is a way in which men exercise control

over women (Dobash and Dobash, 1979; Eswaran and Malthotra, 2011; Martin, 1976;

Wilson and Daly, 1993, 1996; Yllo and Strauss, 1990). A number of studies have reinforced

the notion of Ghana as a patriarchal country. Ofei-Aboagye (1994), for example, described

how in proverbs, oral traditions and folklore, women were celebrated for their ‘life-giving

capacities’. The duty of husbands, on the other hand, was to secure the obedience and

fidelity of the wife. Cultural traditions, such as wife inheritance and dowry payments,

forced marriages, widowhood rites and FGM, have reinforced the belief in men’s

superiority and encouraged violence against women (Adu-Gyamfi, 2014; Amoah, 2007;

Amoakohene, 2004; Adomako Ampofo, 1993; Tenkorang et al., 2013). This is particularly

true when women are dependent on men because of limited education and training

opportunities, and restricted access to cash, credit and employment (Amoakohene, 2004;

UNICEF, 2000). Mann and Takyi (2009) found further that respondents from matrilineal

groups were less likely to condone spousal abuse than those from patrilineal backgrounds

(see also Panda and Agarwal, 2005). This study will investigate further the role of

patriarchal norms as a determinant of domestic violence through the analysis of specific

questions included in the qualitative and quantitative instruments to uncover how

normative values at the community and society levels may affect individual and household

behaviours and attitudes towards domestic violence.

Masculinities. The way men express masculinity in multiple and, sometimes,

contradictory ways has been linked to gender-based violence (Edström et al., 2014a, 2014b;

Freedman and Jacobson, 2012). The focus on masculinities since the late 1980s has been due

to a number of factors with significance for sub-Saharan African countries, including the

emergence of sexually transmitted diseases such as HIV (Barker and Ricardo, 2005; Eves,

2010; Harrison et al., 2006; Mills et al., 2009), resource-scarcity and conflict-related

migration (Campbell, 1997; Jones, 2006; Cornwall and White, 2000; Little and Jones, 2000),

and social and economic crises that trigger crises of masculinities (Richards, 1996, 2005).

The ‘crisis of masculinities’ literature, however, has been recently challenged. Studies in

Sierra Leone, South Africa and Kenya, among other countries, have emphasised the

importance of fostering positive masculinities among men to encourage them to challenge

harmful patriarchal norms and collectively address gender-based violence (Mills et al.,

39

2015a, 2015b). In Ghana, Mann and Takyi's (2009) study found that expectations of

masculinity that focused on men’s financial responsibility to provide for their households

was linked to men believing that intimate partner violence was acceptable.

The pressures placed on men, through these and other harmful constructions of

masculinity, not only reinforce harmful gender dynamics and violence but have also been

found to contribute to men’s ill health. For instance, Adinkrah (2012) investigated male

suicide behaviour in Ghana using data from the Statistics and Information Technology

Unit of the Ghana Police Service. The study found that over half of the 287 (attempted)

suicides covered in the data were motivated by an attempt to avoid the dishonour,

humiliation, public disgrace and shame associated with economic failure, disease and

physical conditions, deviant or criminal conduct and sexual impotence. In a follow-up

study, Adinkrah (2014) focused on homicide-suicides. The majority of the 50 cases

analysed between 1990 and 2009 showed a particularly severe form of intimate partner

violence: they involved husbands suspecting infidelity and men whose wives threatened

them with divorce or separation. Studies on masculinities, both in and beyond Ghana,

therefore, show the harmful effects of negative masculinities on both men and women.

Emerging studies have also revealed that, as a socially constructed phenomenon,

masculinities can be positively harnessed to transform detrimental gender dynamics (Mills

et al., 2015a, 2015b).

Exposure to other forms of violence. A smaller literature has examined how domestic

violence may be related to other forms of violence prevalent within the wider society.

Recent evidence has shown that exposure to a violent conflict or civil war at some point of

an individual’s life could be associated with rises in domestic violence. La Mattina (2012)

found that exposure to the 1994 Rwandan genocide led to an increase in domestic violence

against women who married after the genocide. Gallegos and Gutierrez (2011) found a

positive and large effect of conflict exposure in Peru on the incidence of domestic violence,

particularly when conflict exposure took place during late childhood and teenage years.

The study reported, in addition, that women who were exposed to civil conflict at a

younger age were more likely to justify wife-beating. Noe and Rieckmann (2013) showed

that women living in districts in Colombia with high conflict intensity had a higher

probability of being a victim of domestic violence than those living in districts with low

conflict intensity. These results may be due to post-traumatic stress disorder or

desensitisation caused by exposure to wider violence. This literature is only in its infancy.

This study has attempted to analyse further this interesting hypothesis through questions

in the qualitative and quantitative instruments on experiences of violence, crime and safety

in and outside the communities sampled. Results are discussed in Section 4.

2.4. Consequences of domestic violence

The persistence of domestic violence is not only a violation of basic human rights but also

restricts the achievement of gender equality across a variety of social, economic and

political outcomes. Research on the consequences of domestic violence is, however,

40

limited. The few studies that have analysed the consequences of domestic violence show

that domestic violence is typically associated with a long list of adverse consequences –

notably on physical and mental health, and on the capacity of women and men exposed to

violence to carry on with their work and other daily activities. The literature has also

reported large direct and indirect effects of domestic violence within the household on

children’s physical and emotional well-being and future life prospects.

Physical and mental health. The most widely reported consequences of domestic violence

are poor physical and mental health. Adverse effects include physical injuries, pregnancy

complications, depression and higher risk of contracting HIV and other sexually

transmitted diseases (García-Moreno et al., 2005; Moosa, 2012). García-Moreno et al. (2013)

reported that women who had experienced physical or sexual abuse by their partners were

16 per cent more likely to give birth to low-weight babies than women who did not

experience intimate partner violence. They were also almost twice as likely to experience

depression, up to 1.5 times more likely to contract HIV, and more than twice as likely to

have an abortion. García-Moreno et al. (2005) showed that domestic violence has

particularly severe consequences on mental health: women who had ever experienced

physical or sexual violence by an intimate partner reported significantly higher levels of

emotional distress,13 and were more likely to have thought about suicide or have attempted

to commit suicide. In the case of Ghana, Amoakohene (2004) reported widespread adverse

psychological and emotional effects as a result of domestic violence: women in the study

mentioned being in constant fear when their husbands came home, depressed over their

lack of freedom, and suffering from stress, tension and low self-esteem. Three other recent

studies from Ghana showed harmful effects of domestic violence on taking pregnancy to

full term and the health of babies born (Pool et al., 2014), between domestic violence and

health status (Addai and Adjei, 2014), and between domestic violence and mental health

(Sipsma et al., 2013). Adu-Gyamfi (2014) reported similar non-physical side effects of

domestic violence, such as an erosion of a person’s sense of dignity, self-esteem, confidence

and respect, particularly when injuries make it difficult for victims to go out in public for

fear of being ‘found out’.

Work and daily activities. Some studies have noted a close association between

experiences of domestic violence and restricted access to livelihood options, fewer skills

and lower productivity among victims (García-Moreno et al., 2005; Moosa, 2012). Lloyd

(1997) found that female victims of domestic violence in the USA were more likely to have

been unemployed, had multiple jobs, had lower personal incomes and were significantly

more likely to receive public assistance than women who did not experience domestic

violence. Similar results were found in Friedman and Couper (1987) and Shepard and

Pence (1988), also for the USA. In Colombia, Ribero and Sanchez (2005) showed that

domestic violence against women caused increased unemployment (by 6.4 percentage

points), a 40 per cent reduction in earnings, and adverse health outcomes. In Ghana,

Britwum and Cusack (2009) found that individuals who experienced domestic violence

13 Emotional distress was identified through symptoms such as crying easily, inability to enjoy life, fatigue,

and thoughts of suicide in the four weeks prior to the interview (García-Moreno et al., 2005).

41

took time off work as a direct effect of the violence.14 Duvvury, Minh and Carney (2012)

showed that domestic violence also affects men, with men losing about 6.5 days of work

per incident of violence. Eswaran and Malthotra (2011) found a negative effect of domestic

violence on women’s autonomy, while Bloch and Rao (2002) showed that domestic

violence was a means used by perpetrators (males in their study) in India to ensure the

maintenance of their bargaining power within the household.

Child development and behaviour. Domestic violence within families is associated with

direct and indirect effects on children. The most prominent consequences include lower

birth weight, lower IQ scores and increased emotional and behavioural problems (Aizer,

2011; Koenen et al., 2003; Sternberg et al., 1993; Wolfe et al., 2003). Carrell and Hoekstra

(2010) have also shown that children who suffered from social and emotional problems

related to exposure to domestic violence also experienced lower academic achievement.

To capture the consequences of domestic violence for individuals, their families and their

communities, the quantitative survey conducted in this study included questions about the

immediate consequences of domestic violence, particularly physical and sexual violence.

Questions elicited information about whether and how violence influenced men and

women’s ability to work, attend school or provide childcare, as well as about health and

psychological consequences, including fear and confidence. This information was

complemented by a number of interviews and focus group discussions designed to

understand the consequences of domestic violence at the level of the individual, the

household and the community. Specific questions addressed social relations, the well-

being of children, and levels of cohesion within households and communities, including

tensions that may result from broken marriages and the escalation of conflicts beyond the

immediate relationship between victim and perpetrator.

2.5. Use of and access to victims’ support services and institutions

The final research question addressed by this study asks how men and women affected by

domestic violence use and access support services and institutions. Levels of domestic

violence are typically under-reported across the world. García-Moreno et al. (2005) asked

about the use of formal services across several countries (health services, legal advice and

shelters), and whether victims of domestic violence asked for support from the police,

women’s NGOs, local leaders or religious leaders. The majority of physically abused

women reported that they had never approached any formal organisation or service. When

asked about the reasons for (not) seeking help, most respondents mentioned a general

acceptance of these behaviours as normal. Others feared consequences such as further

violence, losing their children or bringing shame to their family. Seeking help – a limited

occurrence in the sample – was usually associated with the severity of the violence, the

14 This study interviewed 1,588 women aged over 19 years and 481 girls aged 13–18 years. Nearly a third

reported having experienced physical violence, a quarter had experienced psychological violence, and one

fifth had experienced sexual violence at least once in their lives.

42

potential negative impact on children and encouragement from friends and family. Only

victims of very severe acts of violence left their husbands.

Similarly, Amoakohene (2004) found that victims of domestic violence in Ghana generally

did not report abuse. Support was not always forthcoming when the victims approached

relatives, the village chief or the police. The main explanation for these findings was the

widespread view that domestic violence was normal. Other barriers that have prevented

victims of domestic violence in Ghana from seeking help from outside the family included

concern for the children’s economic and emotional well-being, lack of family support, lack

of information or economic means to pursue legal rights and the belief that women have

to obey their husbands (Coker-Appiah and Cusack, 1999; Ofei-Aboagye, 1994).

Several institutions have been set up to respond to domestic violence in Ghana but are

constrained by capacity and other logistical challenges (Mitchell, 2011). Challenges include

the lack of initial and continuing training of staff, the slow pace of the courts and the lack

of specific support services, such as counselling, emergency shelters, legal advice and

medical services. Adu-Gyamfi (2014) analysed the process of prosecuting cases of domestic

violence after the implementation of the 2007 Domestic Violence Act in the Mampong

municipality of Ghana. The study found that the number of abuse reports increased

significantly after the implementation of the Act but was not accompanied by increases in

the number of prosecutions or convictions. According to the DOVVSU 2011 annual report,

12,706 cases of domestic violence were filed in 2010. Among these, only 954 cases were sent

to court, resulting in only 118 convictions.

One of the objectives of this study was to analyse how resources and services were used in

Ghana to support victims of domestic violence, and how and when victims accessed them.

There appears to be a consensus among service users about a lack of resources and poor

implementation of relevant policies. The perceptions and experiences of authorities who

work at these specialised services are less clear. This study designed a series of structured

observations of these services (discussed in more detail in Section 3), to better understand

the opportunities and barriers for domestic violence survivors to seek and receive support.

To operationalise this framework, the study interviewed several individuals at different

levels, including opinion leaders, activists, legal experts, community leaders (religious and

political leaders, including traditional authorities), legal practitioners and journalists.

The evidence above also suggests that under-reporting of domestic violence may be due

to social normative pressures. This evidence connects to a growing body of feminist

criminology literature that challenges the focus on the criminal justice system as the main

avenue for punishing perpetrators of domestic violence (Buzawa and Buzawa, 2003; Rowe,

2015). These studies show that the apprehension and punishment of perpetrators through

penal and criminal systems might offer short-term protection for women but does not

effectively challenge patterns of violence at a societal or systemic level (Belknap, 2014;

Hoyle, 1998; Sherman et al., 1992). Recent studies show, too, that victims of domestic

violence, especially intimate partner violence, might forgo accessing the criminal justice

43

system altogether because their adjudication wishes (such as anonymity) are not met (Artz,

2014; Cerulli et al., 2015). Further, stigmatisation of those who report violence has also been

found to discourage victims of domestic violence from accessing resources available

through the criminal justice system (Sulak et al., 2014).

The study will explore these factors, alongside other potential explanations such as people

preferring other forms of intervention and mediation due to differing perceptions and trust

about the justice system and other government-appointed support structures (lawyers,

nurses, social workers and teachers, for instance), community leaders and the police. This

approach connects to a current debate in the development literature on whether or not

there is value in using ‘organic’ or ‘existing’ informal mechanisms to mediate individual-,

household- and community-level conflict (Lugo, 2015; Rowe, 2015). Research in Rwanda

and Uganda, for instance, has suggested that ‘organic’ responses foster greater levels of

reconciliation among perpetrators and victims than imposed criminal justice routes

(Doughty, 2014; Burgess and Campbell, 2015). This study will explore how men and

women perceive government institutions and actors (such as police officers and health care

providers). It will also explore how the police and other service providers view the help-

seeking behaviour of citizens. This analysis will make use of a number of questions

included in the formal survey about access to formal and informal services and support

structures, use of these structures and perceptions about their usefulness and effectiveness.

44

3. Methodology

Ghana is a culturally diverse country, and the incidence, causes and consequences of

domestic violence are likely to vary across regions, broader kinship and socio-economic

relations and individual characteristics. The study, therefore, adopted a mixed-methods

approach to collect information about the incidence, nature, determinants and

consequences of domestic violence. This approach involved the collection of primary

quantitative data through a representative household-level survey and qualitative in-

depth data on key mechanisms that may explain trends, patterns and the distribution of

domestic violence in Ghana. The qualitative and quantitative instruments were designed

to capture information related to the dimensions of the social ecological model – personal

histories; individual and household circumstances and relations within; individual and

community perceptions and realities of violence; access to, use of and quality of local

resources and services at local and regional levels; and experiences of violence perpetrated

by a range of possible actors.

From the outset, the aim of the research was to ensure complementarity between the

qualitative and quantitative research methods. The quantitative data enabled the

production of statistics that are regionally and nationally representative at the household

level. The qualitative data explored the perceptions, attitudes and beliefs of women and

men within their everyday lives, to understand how they are related to experiences of

domestic violence as survivors and victims and to observe how individual experiences of

domestic violence are integrated within family, community and wider social and political

structures and relations. The following sections describe the design of the quantitative

survey (subsection 3.1) and the qualitative instruments (subsection 3.2); sampling

procedures for the quantitative survey (subsection 3.3) and qualitative fieldwork

(subsection 3.4); the training of enumerators and pre-testing processes (subsection 3.5); the

implementation of fieldwork across all sites in Ghana (subsection 3.6); response rates

(subsection 3.7); the methods used in the analysis of the qualitative and quantitative data

(subsection 3.8); and the strict ethical procedures followed in the study (subsection 3.9).

3.1. Quantitative instruments

The main quantitative instrument used in the study was a survey entitled the ‘Ghana

Family Life and Health Survey 2015’ (GFLHS 2015). The survey deliberately did not

include any mention of ‘domestic violence’ in the title, and was described to communities

in generic terms as a survey of health and family life. Specific information about the survey

was provided only when the research team had selected the participant within each

household approached, and privacy was ensured. At that point, the survey was explained

in detail to the participant, who was offered the option of refusing to continue with the

interview. Few participants refused to continue (see subsection 3.3 below). This approach

ensured high rates of response because participants were not prevented from answering

the questionnaire by other household members worried about a survey on domestic

violence. The introduction of the study as a health and family life survey also provided a

45

level of safety to the respondents and interviewers, by avoiding any immediate conflicts

or tensions that may have arisen if the topic of domestic violence had been introduced

publicly at the outset, as well as preventing the possibility of adverse follow-up

consequences for the respondent.

The GFLHS 2015 included one household- and one individual-level questionnaire. The

design of the questionnaires was based on the World Health Organization (WHO) survey

for the Study on Women’s Health and Violence against Women (Garcia-Moreno et al.,

2005), DHS and Multiple Indicator Cluster Survey (MICS) domestic violence modules

applied in several countries (see Appendix A), the 1998 study on Violence against Women

and Children in Ghana (Coker-Appiah and Cusack, 1999) and modules used in surveys

implemented by research team members elsewhere. The questionnaire design also built on

the definition of domestic violence used in the 2007 Domestic Violence Act but included

an additional category of domestic violence to capture instances of controlling behaviour

(social violence).

As discussed in the previous section, the definition of domestic violence adopted in this

study is broader than definitions used in standard domestic violence surveys implemented

to date. The questionnaires used in the GFLHS 2015 included questions that address this

wider definition of domestic violence and, therefore, may not be directly comparable to

existing surveys such as the DHS 2008 and the 1999 Gender Centre study (Coker-Appiah

and Cusack, 1999). For instance, the DHS 2008 focused only on physical, sexual and

psychological violence. Therefore, the total incidence of domestic violence captured in this

study will not be directly comparable to the DHS findings. However, by adopting this

broader definition of domestic violence and capturing several acts of violence committed

within domestic relations but not captured before, this study provides for the first time a

comprehensive analysis of how women and men in Ghana experience all forms of domestic

violence, which can be used to fine-tune policy, programmes and monitor future progress.

The household-level questionnaire was designed to capture information about all

individuals in each sampled household. For the purpose of the study, the standard

definition of a household used by the GSS was followed, which includes a group of

individuals who normally live under one dwelling, and prepare and eat their meals

together. The household survey included standard socio-economic questions about age,

sex, education levels, children’s school attendance, occupation, work activities and related

payments, religion and ethnicity. The household questionnaire also collected information

on food consumption, household responses to economic shocks, asset ownership, decision-

making processes within the household, and conflicts in the community.

The individual-level questionnaire was administered to only one man or one woman in

each sampled household, in line with WHO (2001) guidelines. This questionnaire included

five modules on individual experiences of domestic violence from the perspective of both

victims and perpetrators. The modules followed the definitions and typologies of domestic

violence discussed in the previous section, and elicited information about experiences of

46

social, sexual, physical, psychological and economic violence. These modules were applied

to all respondents and woven into the questionnaire between other modules to avoid

repetition and respondent fatigue. Other additional modules included questions designed

to capture the various dynamics, determinants and consequences of domestic violence

across the four spheres of the social ecological model: safety in the community, social

capital, relationships within and outside the household and local community,

reproductive, physical and mental health, individual agency, access and use of services,

and approaches to parental and child discipline. The individual questionnaire also

contained a detailed module on attitudes towards domestic violence that made use of

attitudinal questions, scenario questions and vignettes, as used, for instance, in the DHS

domestic violence modules (García-Moreno et al., 2005; see also Thaler, 2012). The

complete individual and household questionnaires are included in the Questionnaire

Annex at the end of this report.

3.2. Qualitative instruments

The qualitative instruments included community-level focus group discussions,

interviews with community leaders and key informants, field observations and individual

in-depth interviews. The complete list of instruments used in the qualitative stage of

fieldwork is provided in the Questionnaire Annex at the end of this report.

The qualitative instruments were designed to generate data that could be analysed across

sex, age and location (rural/urban), and to triangulate key findings across the different

social ecological spheres. Focus group discussions generated data at the individual, family

and community levels and explored interactions across each of these levels. Interviews

engaged closely with community leaders and service providers to generate data at the level

of the community and the broader social and political contexts.

The focus group discussions took place among four distinct groups: (i) men between 18

and 50 years old who were married or living with a partner; (ii) women between 18 and 50

years old who were married or living with a partner; (iii) men and women between 18 and

25 years old who were unmarried; and (iv) men and women over 50 years old who were

married or unmarried.

The groups were disaggregated by sex and age to allow a better understanding of how

these factors may affect different responses to the research questions. Group sizes varied

between 8 and 12 people, and were guided by one facilitator and supported by one note-

taker and one translator. Further details about the composition of the groups are provided

in Appendix C. The discussions were conducted in the main local language of each region

and explored norms, beliefs and practices regarding domestic violence, as well as how and

when individuals affected by domestic violence seek medical and legal advice. Exact

questions are provided in the Questionnaire Annex at the end of this report.

47

The focus group discussions were able to draw out a great deal of detail on perceptions,

determinants and consequences of and responses to domestic violence at the level of

individuals, households, communities and the wider society. The questions not only

probed interactions across the model’s spheres and the different contexts in which domestic

violence takes place (such as asking about violence in public arenas such as market places,

compared to violence in private spaces such as homes), but they also explored different

types of domestic violence (controlling behaviour, economic violence, sexual violence and

so forth). Because the respondents were clustered in specific groups, it was also possible to

understand some of the factors (for instance, sex, age, location) that explain the incidence

of domestic violence, compared to others.

The qualitative instruments also included in-depth interviews with community leaders

and key informants, designed to understand experiences of violence and perceptions about

the causes and consequences of domestic violence, and ways and barriers to address it.

These interviews provided important insights into the practicalities of everyday work

related to domestic violence, particularly with respect to law enforcement and engagement

with community organisations dealing with domestic violence. By conducting interviews

with opinion leaders alongside the focus group discussions, it was possible to generate a

detailed picture of how people access state resources (such as health care services),

alongside the factors that affect the capacity of service providers to meet the needs of those

affected by domestic violence. These interviews enabled the team to also explore the

interplay of dynamics occurring at a political and economic level (for example, the failure

to provide resources to remote clinics or police stations), and the implications they have

for community workers, families and individuals who are affected by domestic violence.

Details about the nature and number of interviews are included in Appendix C.

The qualitative fieldwork also made use of observations made in the main health facilities

and police services of each selected district in each of the ten regions. The aim of these

observations was to examine first-hand the availability and use of public services, and

understand the barriers that women, men and service providers face. These observations

made it easier to understand the context and situations that people described in the in-

depth interviews and focus group discussions. Details of the observations conducted in the

study are provided in the Questionnaire Annex at the end of this report. The observations

were used in particular to complement individual interviews with key staff at the observed

institutions. These explored the barriers officials face in implementing national-level

policies, and their views about the relevance and effectiveness of policies to address

domestic violence.

In addition to the methods described above, 15 in-depth individual interviews were

conducted with women and men affected by domestic violence identified through specific

trigger questions in the quantitative survey. These interviews took place privately

immediately after the individual surveys by enumerators specifically trained on the use of

qualitative data collection methods (see subsection 3.3 below). These interviews allowed

more time with the respondents to discuss in detail the occurrence, perceptions,

48

consequences and ways of coping with domestic violence. Although at first the survey

team was concerned about potential respondent fatigue, it found instead that respondents

seemed “excited and appreciative to get someone to share their story with”.15 Detailed

questions are included in the Questionnaire Annex at the end of this report.

3.3. Sampling design: quantitative survey

The study focused on men, women, boys and girls between the ages of 15 and 60 years.

This age group extends the usual age group included in international surveys (typically 15

to 49 years), to capture the incidence of domestic violence among older women and men,

an issue brought to the attention of the research team during initial scoping interviews

with key stakeholders in Ghana. Children below the age of 15 years old were outside the

scope of the study due to ethical considerations and the fact that a focus on young children

would require a different research design from the one used to elicit information from

adults. The quantitative survey aimed to conduct interviews among representative

samples of individuals aged 15 to 60 years old at national, regional and urban/rural levels,

and by sex. Either one man or one woman was interviewed in each household.

The first stage of the sampling design involved compiling basic information about the

population of Ghana. According to the 2010 census, the population is spread unevenly

across the ten regions of the country, with each region having between 2.5 and 19.4 per

cent of the overall population (see Appendix D). The level of urbanisation also varies

significantly between regions, ranging between 16.3 and 90.5 per cent.16 Sex ratios are on

average balanced, varying only slightly across regions. This information was used to

stratify the population of Ghana into 20 strata at the regional level, and by rural and urban

areas within regions. The sample design then followed a two-stage procedure. The first

stage consisted of randomly selecting primary sampling units within the 20 strata using

the sampling frame of the 2010 census enumeration areas (EAs, also referred to as

‘clusters’). In the second stage, 15 households were randomly selected from a listing done

by the GSS team in an average of 30 EAs in each region.

When designing a representative sample, it is important to take into account the prevalence

of the outcome being measured (domestic violence, in the case of this study). According to

GSS et al. (2009), the prevalence of domestic violence in 2008 varied considerably between

regions (between 26.8 and 49.4 per cent for women, and between 14.3 and 42.3 per cent for

men). 17 To be confident that the GFLHS 2015 would capture the ‘true’ prevalence of

domestic violence, the study used a more conservative estimate of 50 per cent incidence

across both sexes.

15 Correspondence with Mr Peter Takyi Peprah (Project Coordinator), 3rd February 2016. 16 Urban localities were defined as localities with 5,000 inhabitants or more (Ghana Census, 2010). 17 ‘Domestic violence’ in the DHS 2008 is defined as physical, sexual or psychological violence by

husband/wife/partner against the respondent.

49

The sample size per region required to accurately estimate a prevalence of 50 per cent

across the population of Ghana is given by the following formula (Daniel, 1999):

n = 𝑍2(𝑝(1−𝑝))

𝑑2× 𝐷𝐸,

where n = sample size, z is the z statistic for a level of confidence, p is the prevalence, d is

the level of precision chosen, and DE represents the design effect. Originally, a statistical

confidence of 95 per cent was chosen, which is associated with a Z-score of 1.96. The

desired precision of the estimated prevalence was set at 5.5 per cent.18 The design effect

(DE) was calculated as follows:

𝐷𝐸 = 1 + 𝜌(𝑚 − 1),

where 𝜌 is the intra-regional correlation coefficient, and m the number of regions (10).

Similar to García-Moreno et al. (2005), the intra-regional correlation for the different types

of domestic violence in the Ghana DHS 2008 data is very low, usually around 1–2 per cent,

and never above 4 per cent. As a result, a conservative intra-regional coefficient of 7 per

cent was chosen. These calculations resulted in a sample size of 4,995 individuals in 331

EAs. Appendix D shows the allocation of the EAs for each region and describes the survey

weights.

3.4. Sampling design: qualitative fieldwork

The design of the qualitative fieldwork was based on information on the length of the

various instruments, travel times, time needed in each community, transcript preparation

times and time spent on the analysis of the data. To minimise travel costs and times, the

team selected one district in each of the ten regions of Ghana. Since in all ten regions the

regional capital covers the full district and there is no rural site, the district with the second

largest urban town within the region was selected. This ensured that both urban and rural

sites were covered within the same district. The second most populated urban town was

selected as the urban field site, and the least populated in that same district was selected

as the rural study site.

Overall, the team conducted 80 focus group discussions with women and men across

Ghana, 210 in-depth interviews with opinion leaders, activists and legal experts, and

observations and in-depth interviews with institutions of justice (20) and health (30). The

selection of participants for the qualitative fieldwork differed depending on the qualitative

tool used and type of respondent. For focus group discussions with community members

within urban areas, respondents from each peri-urban, lower-, middle- and upper-class

areas were selected using the help of local assembly staff to map out residential areas by

18 The precision parameter is related to the confidence interval of the estimate, with the width of the

confidence interval being 2d. We increased slightly this precision from the conventional 5 per cent to 5.5 per

cent due to budgetary constraints.

50

class. In rural areas, participants for the focus group discussions were randomly selected

using a list of households in each of the areas, obtained from GSS.

For interviews in and observations of police facilities, a representative at the district police

office was chosen, as well as staff of the DOVVSU unit at the district police office – if one

existed. If the district police office did not have a DOVVSU unit, the DOVVSU unit that

covered the area of the district police office was selected instead. There, the research team

interviewed the head of the unit and one assistant, and conducted on-site observations of

its services.

With respect to health facilities, the research team selected in each district two or three

members of staff from the district hospital’s outpatient department and emergency unit

and, if available, counsellors or psychologists working with individuals affected by

domestic violence. Many sampled districts in rural areas (which covered the least

populated rural area) did not have health facilities. Instead, staff of available community-

based health planning services were interviewed as part of the interviews with opinion

leaders. Exact details are provided in Appendix C.

In both urban and rural areas, traditional chiefs, the queen mother,19 a religious leader, a

headmaster or a teacher of a public school, a traditional healer and/or a private health

practitioner were selected to participate in the community/opinion leader interviews. In

urban areas, the person responsible for gender-related services at the district assembly was

also approached. Depending on availability, one or two representatives from each category

of those listed were chosen in each site. Certain categories of respondents, such as legal

practitioners and other professionals, were not resident in rural areas and were, therefore,

selected in urban sites only. All sub-groups of opinion leaders in each community were

found and interviewed (see Appendix C). Exceptions were health care workers in the

Upper East and Volta regions, where two (and not the planned three) respondents were

available to interview. These were replaced by available representatives of other eligible

groups of opinion leaders. Additional key informant interviews in each district targeted

legal practitioners identified by the Commission on Human Rights and Administrative

Justice (CHRAJ), activists from organisations working on domestic violence at national and

regional levels identified by the DV Coalition, media professionals (radio or TV) in each

district, and journalists at the national level.

3.5. Training and pre-testing

The implementation of all quantitative and qualitative instruments was preceded by

extensive training, and by in-depth pre-testing of all instruments.

19 Queen mothers are women who rule alongside the chiefs of a traditional area and who play a central role

in the traditional governance of that area. They typically come from a royal line of descendants within a

particular clan/family, usually the same as that of the traditional chief. In Ghana, queen mothers are more

prominent in the Akan culture, which practises matrilineal descent. Queen mothers’ traditional intermediary

roles have been used by international networks and local groups campaigning against domestic violence.

51

The training of the quantitative fieldwork team was done in two stages. The first training

event took place in early May 2015 using draft questionnaires. The main aim of this training

was to prepare the pre-testing of the survey instruments. The training consisted of six main

components: (i) introduction to the study and sensitisation about domestic violence in

Ghana; (ii) outline of the quantitative survey and how to introduce oneself and the study

to households; (iii) familiarisation with the questionnaires; (iv) introduction to the use of

tablet PCs and related software; (v) working through examples of the questionnaires using

tablets; (vi) discussion about research ethics; and (v) practice in the field, followed by a

detailed debrief session. The survey questionnaires were pre-tested after this training

exercise to assess the feasibility and structure of the questions, improve the flow of the

questionnaire and refine answer codes.

The re-drafted final questionnaires were then used as the basis for the second training

exercise in late May 2015. This session offered the team the opportunity to refine the

questions and answer codes further, and to adjust how questions were phrased.

Enumerators worked in different language groups to agree on the language and

terminology they would use when translating certain terms in the questionnaire. This was

important to ensure that all research methods used the right language and measured real

incidence levels and perceptions of domestic violence. The second training event also

included special exercises on qualitative methods involving members of the quantitative

survey team assigned to conduct the in-depth individual interviews.

The training of the qualitative fieldwork team took place in early April 2015. Eighteen

qualitative researchers affiliated with ISSER and 18 enumerators from GSS took part in this

event. Supervisors of the three teams (of five members each) were selected on the basis of

their leadership expertise in past research projects. The topics covered by this training

exercise included: (i) violence, gender and health, including background information on

domestic violence in Ghana; (ii) overview of the methods used for the data collection; (iii)

discussion about research ethics; (iv) training on the use of in-depth interviews and focus

group discussions, including selection criteria, planning and implementation; (v) theory,

practice and quality control with respect to notes and transcripts and to observational

methods; (vi) methods for data analysis; and (vii) planning fieldwork. The training event

was based on role-playing and mock interviews, in which scenarios were played out to

help identify the ‘do’s and don’ts’ of sensitive behaviour towards participants, and to test

the logic, length and practicality of the drafted instruments. The training of the qualitative

field team was also used as an opportunity to refine the instruments, which were tested in

the field during a piloting exercise immediately after the training event.

3.6. Fieldwork implementation

Fieldwork took place between April and August 2015. The GFLHS 2015 data collection

started on 5th June and lasted until 26th August. Fifteen teams of three enumerators and one

supervisor visited 331 EAs overall. The teams spent 2.5 days in each EA, devoting one day

52

to the listing of all households and 1.5 days to conducting 15 interviews. The quantitative

data collection was done using tablet PCs, which allowed for greater efficiency in terms of

time spent in each EA, and better precision of the answers entered into the questionnaire.20

The use of tablets also allowed the teams to send data regularly to a main server. Incoming

data were then observed through graphs and exported to a spreadsheet, which allowed

the team to look out for errors and inconsistencies in almost real time.

The qualitative fieldwork took place between 27th April and 26th July. The qualitative teams

travelled and collected data for two weeks and then returned for one week to Accra to

finalise their transcripts. They repeated this process four times. This structure was

designed to reduce pressure on the field team, and increase the quality of the transcripts.

The time spent in Accra also allowed the qualitative research lead to perform a second

round of quality checks.

Control systems were put in place to ensure quantitative and qualitative data quality. The

survey data were checked twice – first by the supervisors before uploading the data to the

server, and then by a data monitoring team, who listed and plotted key variables to look

for outliers and inconsistencies. Any abnormal findings were fed back to and addressed

by the field supervisors. The supervisors also visited the field teams regularly to carry out

quality control checks and provide technical and emotional support. With respect to the

qualitative fieldwork, interviewers peer-reviewed each other’s notes every day before

filing them. Transcripts were reviewed several times during the weeks spent in Accra.

Spending several days in any one area also allowed the qualitative team to identify gaps

and inconsistencies in time and to go back to the respondents if required.

3.7. Response rates

Table 1 shows the survey response rates. These rates ranged from 97.6 per cent in the

Western region to 99.6 per cent in Brong Ahafo.21 These high rates of response were due to

the precautions taken when administrating the survey. First, as discussed above,

information about the survey being about domestic violence was only disclosed after the

individual participant had been chosen and privacy had been ensured. Second, the

household questionnaire was designed so that it could be administered to heads of

households without disclosing the content of the individual questionnaire. This was done

because past experience of the GSS with household surveys in Ghana had shown that an

engagement with heads of households – usually male – would subsequently lead to higher

participation and completion rates among other household members. Third, the individual

questionnaire – containing the modules on experience and perpetration of domestic

violence – was administered to a randomly chosen individual aged 15–60 years using tablet

20 For example, data are only entered once, rather than recorded on paper and then entered electronically,

which minimises data entry error. 21 The percentage of completed interviews is the same as the household response rate, as there were no vacant

or destroyed dwellings or absent households, thanks to the listing exercise that preceded the survey

collection.

53

PCs. The tablets in turn enabled the research team to make use of pictorial responses, as in

García-Moreno et al. (2005), where respondents were able to choose a ‘X’, depicting no

experience of physical violence or sexual abuse (used for both separately), or a ‘tick’

depicting having been subject to physical or sexual abuse in the past (see the questionnaires

in the Questionnaire Annex at the end of this report). This choice was made at the end of

the physical and sexual violence module, without the enumerator observing it. These tools

were used to counteract potential levels of under-reporting of these types of violence, as

indicated in other studies.

Table 1: Response rates to household survey (in percentages)

W C GA V E A BA N UE UW Total

Completed 97.6 98.1 98.2 98.4 98.3 98.9 99.6 98.8 98.1 97.8 98.5

No one at

home

2.4 1.9 1.4 1.4 1.5 0.9 0.4 0.8 1.9 1.5 1.3

Refused 0.0 0.0 0.1 0.0 0.2 0.0 0.0 0.4 0.0 0.0 0.1

Dwelling

destroyed

0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Source: Ghana Family Life and Health Survey (GFLHS) 2015.

Notes: W: Western region; C: Central region; GA: Greater Accra region; V: Volta region; E: Eastern region; A:

Ashanti region; BA: Brong Ahafo region; N: Northern region; UE: Upper East region; UW: Upper West

region.

The response rates to the individual survey were also consistently high (Tables 2 and 3).

Response rates for women ranged between 96.3 per cent in the Volta region and 100 per

cent in Upper West (Table 2). Response rates for men ranged between 98 per cent in Eastern

region and 100 per cent in the Upper East and Upper West regions (Table 3).

54

Table 2: Response rates to the individual survey: women (in percentages)


Completed 98.9 99.6 98.2 96.3 99.1 99.3 99.0 98.9 99.0 100.0 98.8

Not available 1.1 0.4 1.0 1.6 0.6 0.0 0.7 0.4 1.0 0.0 0.6

First

respondent did

not allow

second

respondent to

answer

0.0 0.0 0.0 0.4 0.0 0.0 0.3 0.0 0.0 0.0 0.1

Second

respondent

refused

0.0 0.0 0.2 1.2 0.3 0.3 0.0 0.4 0.0 0.0 0.3

Lack of privacy 0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.4 0.0 0.0 0.1

Other 0.0 0.0 0.0 0.4 0.0 0.3 0.0 0.0 0.0 0.0 0.1

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0




region.

Table 3: Response rates to the individual survey: men (in percentages)


Completed 99.0 99.3 98.4 98.4 98.0 98.6 98.1 99.2 100.0 100.0 98.7

Not available 1.0 0.0 1.3 0.5 0.5 0.3 0.6 0.4 0.0 0.0 0.6

First

respondent did

not allow

second

respondent to

answer

0.0 0.0 0.0 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.1

Second

respondent

refused

0.0 0.0 0.0 0.5 0.0 0.3 1.2 0.0 0.0 0.0 0.2

Lack of privacy 0.0 0.0 0.3 0.0 1.5 0.3 0.0 0.4 0.0 0.0 0.3

Other 0.0 0.7 0.0 0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.2

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0




region.

Table 4 presents the breakdown of the sample along key socio-demographic characteristics.

Just over a quarter (27 per cent) of the individuals sampled were aged 30–39 years. Only 8

per cent were aged 15–19 years. The majority of respondents (60 per cent) were self-

employed, and 17 per cent were employed. Over two thirds of the sample were married or

living with someone. One fifth of the individuals sampled were divorced, separated or

55

widowed, and one tenth had never been married. The most common education level was

middle school (37.7 per cent), followed by no education (19 per cent), secondary and

primary education (around 15 per cent each), higher education (9.5 per cent) and technical

school (3 per cent). In terms of asset quintiles,22 20 per cent of all individuals in the sample

are in the bottom quintile, 26.5 per cent in the second quintile, 16.6 per cent in the third

quintile, 21.2 per cent in the fourth quintile, and 15.8 per cent in the top quintile.

Table 4: Distribution of survey respondents by key socio-economic characteristics

(weighted)

Sample All All Women Women Men Men

number percentage Number percentage number percentage

Age group

15–19 394 7.9 191 6.4 203 10.3

20–24 682 13.7 380 12.7 301 15.2

25–29 936 18.8 556 18.6 378 19.1

30–39 1,362 27.4 886 29.6 475 24.0

40–49 856 17.2 512 17.1 343 17.3

50–60 744 15.0 463 15.5 281 14.2

Total 4,974 100.0 2,989 100.0 1,982 100.0

Employment

Self-employed 2,984 59.7 1,951 65.3 1,032 52.0

Employed 845 16.9 322 10.8 522 26.4

Not working 1,166 23.3 716 24.0 428 21.6

Total 4,995 100.0 2,989 100.0 1,982 100.0

Marital status

Never married 496 10.0 156 5.2 339 17.1

Married or living together 3,472 69.9 2,189 73.4 1,281 64.7

Divorced, separated or

widowed

998 20.1 638 21.4 359 18.1

Total 4,966 100.0 2,984 100.0 1,979 100.0

Residence

Urban 2,559 51.2 1,580 52.9 964 48.7

Rural 2,436 48.8 1,409 47.1 1,018 51.3

Total 4,995 100.0 2,989 100.0 1,982 100.0

22 Individuals in the lowest asset quintile have four assets from the list provided in the individual

questionnaire (see Questionnaire Annex). Individuals in the second quintile have between five and seven

assets. Individuals in the third quintile have between eight and nine assets. Individuals in the fourth quintile

have between 10 and 12 assets. Individuals in the highest quintile have between 13 and 21 (maximum) assets.

56

All All Women Women Men Men

number percentage Number percentage number percentage

Region

Western 506 10.1 280 9.4 226 11.4

Central 440 8.8 290 9.7 145 7.3

Greater Accra 949 19.0 590 19.7 351 17.7

Volta 454 9.1 251 8.4 200 10.1

Eastern 488 9.8 303 10.1 185 9.3

Ashanti 954 19.1 575 19.2 378 19.1

Brong Ahafo 507 10.1 331 11.1 175 8.8

Northern 411 8.2 216 7.2 192 9.7

Upper East 173 3.5 86 2.9 86 4.4

Upper West 114 2.3 67 2.2 45 2.3

Total 4,995 100.0 2,989 100.0 1,982 100.0

Education level

None 958 19.3 684 22.9 274 13.8

Primary 751 15.1 504 16.9 246 12.4

Middle/JSS/JHS* 1,871 37.7 1,167 39.1 703 35.6

Secondary 769 15.5 361 12.1 407 20.6

Technical 144 2.9 78 2.6 66 3.4

Higher 472 9.5 190 6.4 281 14.2

Total 4,964 100.0 2,984 100.0 1,977 100.0

Asset quintile

Lowest 999 20.0 583 19.5 410 20.7

Second 1,324 26.5 802 26.8 517 26.1

Third 828 16.6 487 16.3 338 17.1

Fourth 1,057 21.2 667 22.3 384 19.4

Highest 788 15.8 450 15.1 332 16.8

Total 4,995 100.0 2,989 100.0 1,982 100.0


Note: Differences in totals are due to missing values in the respective questions.

* JSS = junior secondary school; JHS = junior high school.

The survey captured a good balance of male and female respondents in the household

interview. There are, however, more female respondents (60 per cent) in the individual

interviews because men were much more often away from the household than women – a

common feature of most household surveys. This did not affect the analysis or main

results, which control for sex composition.

The geographical representativeness of the sample was also close to the census

distributions. Table 4 shows the distribution of the sample across rural and urban areas:

48.8 per cent in rural areas (49.1 per cent in the 2010 census) and 51.2 per cent in urban

areas (50.9 per cent in the 2010 census). It also shows the distribution of the sample across

regions: 10.1 per cent in the Western region (9.6 per cent in the 2010 census), 8.8 per cent in

the Central region (8.9 per cent in the 2010 census), 19 per cent in Greater Accra (16.3 per

57

cent in the 2010 census), 9.1 per cent in the Volta region (8.6 per cent in the 2010 census),

9.8 per cent in the Eastern region (10.7 per cent in the 2010 census), 19.1 per cent in the

Ashanti region (19.4 per cent in the 2010 census), 10.1 per cent in the Brong Ahafo region

(9.4 per cent in the 2010 census), 8.2 per cent in the Northern region (10.1 per cent in the

2010 census), 3.5 per cent in the Upper East region (4.2 per cent in the 2010 census) and 2.3

per cent in the Upper West region (2.3 per cent in the 2010 census).

3.8. Data analysis

The quantitative survey data were weighted to account for differences in the selection

probability and in the number of completed interviews among primary sampling units (see

details in Appendix D). The quantitative data were analysed using cross-tabulations

between domestic violence outcomes and socio-economic variables such as age, sex,

region, educational level, marital status, employment status and wealth, among others.

The study also used multivariate regressions to understand in more detail the

determinants and consequences of the various forms of domestic violence.

Regression analysis account for the correlations between potential determinants or

consequences of domestic violence – something that is not possible when using only cross-

tabulations. For instance, individuals with higher levels of education also tend to be

wealthier. It is then possible that both higher levels of education and higher levels of wealth

are associated with higher incidence of domestic violence in isolation but not together. It

may be that, in fact, one of those factors is more important than the other. Alternatively,

both variables may be independent drivers of domestic violence. Multivariate regressions

help to identify the most significant determinants of domestic violence by taking into

account correlations between those determinants.

Second, multivariate regressions enable the analysis of regional differences. Domestic

violence in Ghana varies considerably across regions. This may be because regions are

different in the ways in which domestic violence occurs. However, the same result may

also be explained not by genuine cultural differences between regions but, rather, because

regions with higher levels of domestic violence are also poorer or have higher levels of less

educated populations. Multivariate regressions distinguish between these two

explanations. This information is, in turn, important for designing the most effective

policies.

Third, multivariate regressions allow the inclusion within a single framework of

determinants of violence at the individual, household and wider community or national

level – in keeping with the social ecological framework of domestic violence discussed in

Section 2. It is important to note that while multivariate regressions enable the

identification of the most significant determinants and consequences of domestic violence

in Ghana, the results suggest an association between the variables but do not indicate

causality.

58

The analysis of the qualitative data was based on over 300 transcripts. The data were

analysed using Grounded Theory (Glaser and Strauss, 1998), 23 which encourages an

iterative approach. The qualitative data were analysed using a three-fold analytical system.

First, the data were clustered by region. Two researchers each reviewed data from three

regions, and one researcher reviewed data from four regions. Data were then coded against

the main research questions. This entailed systematically combing through the transcripts

from each region and looking for evidence that related to each of the main research

questions. This first stage of analysis was collated, and circulated to other research team

members to identify any initial queries or inconsistencies. In the second stage, the research

team reviewed one another’s coding, and suggested areas that required further

substantiation or clarification. In the third stage of the process, the researchers generated a

narrative analysis for each research question, based on the empirical material. This analysis

was circulated again among the research team for validation.

Through these three stages of coding, the research team sought to ensure that the claims

made in one set of transcripts, in one region, were validated by triangulating the findings

from these transcripts with data from other regions. During the coding, the following rules

were applied to ascertain validity across a scale of representativeness. ‘Nationally

representative’ findings are those findings that are corroborated by data in all regions, in

both rural and urban sites, and in data from a range of stakeholders (in the key informant

interviews) and group compositions of the focus group discussions. ‘Regionally

representative’ findings are those findings that are supported by data in both rural and

urban sites in one region, and in data from a range of stakeholders (in the in-depth

interviews) and group compositions of the focus group discussions.

3.9. Ethical considerations

The sensitivity of the research topic raised serious ethical and methodological challenges.

Notably, the study involved the collection of data that could have potentially affected the

privacy of subjects, and may have induced some level of stress and anxiety. The study also

involved potential safety risks for interviewers and respondents. Accordingly, the study

followed closely the Ethical and Safety Guidelines developed by the WHO (2001) for

conducting research on domestic violence.

Research was conducted professionally and ethically, with strict respect for principles of

integrity, honesty, confidentiality, voluntary participation, impartiality and the avoidance

of personal risk. Adherence to these principles was overseen and monitored by the team

supervisor and field monitors, in collaboration with the IDS Research Ethics Committee.

The detailed mechanisms that were put in place to ensure compliance with the principles

included:

23 Grounded Theory is a systematic methodology that groups concepts and categories through induction

based on data collected.

59

1. The safety of respondents and the research team was paramount, and guided all decisions

in the study. To ensure the safety of both respondents and the research team, the survey

was introduced to communities and households as a study on women’s health and family

life, as discussed earlier. Interviews were conducted in private settings at convenient and

safe times for the participant, and interviewers were trained to change the subject to a less

sensitive topic if an interview was interrupted. The survey participants were informed

about this beforehand. The safety of interviewers was also ensured by planning visits to

research sites in teams, particularly during the evening and to areas known to be unsafe.

2. Protecting confidentiality was essential to ensure both safety and data quality. The

questionnaires were implemented and analysed anonymously, with the names of

individuals in each household replaced by a numeric code (identifier). Respondents were

informed about protocols for confidentiality and anonymity. No names or addresses

appeared on the questionnaires or transcripts. The survey questionnaires were linked to the

in-depth interview transcripts via the personal identifiers only. The address of individuals

willing to participate in in-depth interviews was noted when they consented to be

interviewed, and destroyed after completion of the interview. Interviewers were trained

and instructed to obtain the permission of respondents to record interviews beforehand

(when this was necessary, for the in-depth interviews and focus group discussions). Tapes

and digital recorders were locked in safe places at all times, and destroyed after

transcription.

3. All research team members were carefully selected and received specialised training and

ongoing support. The training entailed an introduction to gender, gender discrimination

and domestic violence, as discussed earlier. Training also addressed the issue of

stereotypes, biases and fears regarding domestic violence, and provided mechanisms to

overcome these. Field researchers learned and practised how to ask questions in a

supportive and non-judgemental manner, and were given opportunities to come to terms

with emotions evoked during the interview process or by the topic itself (for instance, if the

field researcher had experienced abuse her/himself). Selected field researchers were

encouraged to think about and discuss openly or in private any matters of concern with

training staff and field managers. Field researchers who faced emotional distress during

fieldwork were free to take breaks, carry out less emotionally draining tasks or withdraw

from the fieldwork altogether without judgement. Field researchers were also trained to

regard themselves as researchers and not counsellors. However, they received sufficient

and adequate information to assist those respondents who asked for help (in the form of

pamphlets that directed study participants to appropriate services and counselling).

4. The study design included actions aimed at reducing any possible distress caused to the

participants by the research. As discussed above, the language of the questionnaire and the

behaviour of field researchers was non-judgemental and supportive at all times, and

interviewers were trained on how to respond to distress caused by the interviews. They

were also encouraged and learned how to terminate interviews if necessary. All interviews

60

ended on a positive note, reminding the participants of their self-worth and of the

importance of having shared their experiences.

5. Field researchers were trained to refer individuals requesting assistance to available

local services and sources of support. In cooperation with government agencies and

services, local partners and other stakeholders, such as NGOs and women’s groups, the

research team identified and obtained consent from local formal and informal providers to

direct to them individuals in need. Pamphlets containing discreet information about these

services were given to respondents, regardless of whether they reported being affected by

domestic violence or not.

6. Findings will be properly interpreted and used to advance policy and intervention

development. The research study involved from the outset formal and informal groups

involved in addressing domestic violence, mitigating its consequences and increasing the

well-being of those exposed to domestic violence (see the composition of the Steering

Committee). The research team sought their advice and input at key stages of the study.

The study also paid close attention to the language used when interpreting results, to avoid

stigmatisation or reinforcement of stereotypes of sub-groups. Similar procedures will be

employed when disseminating the results of the study.

61

4. Empirical results This section presents the main findings of the study for each of the research questions

outlined in Section 1. This analysis brings together the qualitative data, and the descriptive

statistics and regression analysis based on the GFLHS. This joint analysis of the qualitative

and quantitative data has allowed the research team to identify points of similarity and

contrast between the data collected using different methods. Each of the subsections below

presents and discusses the main results of the study against each research question in turn.

Subsection 4.1 analyses the incidence of social, physical, sexual, psychological and

economic violence across different population groups in Ghana – independently of

whether this violence was perpetrated or not within domestic relations. Subsection 4.2

focuses on the main remit of the study – the analysis of violence against women and men

perpetrated within domestic relations (i.e. domestic violence). Subsection 4.3 examines the

profiles of perpetrators of domestic violence in Ghana. Subsection 4.4 analyses attitudes

and social norms that drive domestic violence in Ghana across individuals, families,

communities and the wider society. Subsection 4.5 discusses the main determinants of

domestic violence uncovered in the data at the individual, family, community and wider

societal levels. The consequences of domestic violence are analysed in subsection 4.6, with

a focus on physical and mental health and the impact of domestic violence on daily lives

and children. The final subsection (4.7) analyses how women and men affected by domestic

violence in Ghana access and use institutions and available public services.

4.1. Incidence of violence (domestic and non-domestic) against women and men

in Ghana

Conforming to international definitions (e.g. UNICEF, 2002; WHO, 2000), the national 2007

Domestic Violence Act (Act 732), and prior studies conducted in Ghana, the study has

analysed the incidence of the following forms of violence (both within and outside

domestic relations):

social violence – acts of controlling behaviour, such as preventing someone from

seeing friends or family of birth; stopping someone from leaving the house;

requiring to know where someone is at all times; stalking; spreading false

information, videos or photos without permission; or forcing women and girls to

have an abortion;

physical violence – slapping, pushing, shoving, hitting, kicking, dragging or

throwing objects at someone; choking, strangling or burning someone; using a

weapon, hazardous chemicals or substances against someone; or kicking or pulling

someone’s external genitalia (for male respondents only);

sexual violence – acts of unwanted sexual comments or physical contact; rape by

physical force, or otherwise forced sex (for instance, by blackmail or threats); denial

of using protection during sex; a sexual partner hiding their HIV status; sexual acts

or intercourse that were performed on the basis of feeling there was no option; or

penetration with an object against someone’s will;

62

psychological violence – insults, belittling or humiliation in private or in front of

others; threats of abandonment; being ignored or treated indifferently;

intimidations and acts aimed at scaring someone; threats of using weapons against

someone; or threats of hurting someone or someone one cares about; and

economic violence – denial of household money for expenses (chop money) even if

enough financial means are available; unsolicited taking of money; control of

belongings and spending decisions; damage to or destruction of someone’s

property; denial of the right to work; forcing someone to work against their will; or

denial of food and other basic needs.

The GFLHS 2015 asked detailed questions about individual experiences of any of these five

broad categories of violence over the individual’s lifetime, and in the 12 months prior to

the survey. This allowed the study to capture both current incidence levels of overall levels

of (domestic and non-domestic) violence and the lifetime accumulation of experiences of

overall violence among men and women. Table 5 summarises this information.

Table 5: Incidence of violence in Ghana 2015 (in percentages)

Social

violence

Physical

violence

Sexual

violence

Psychological

violence

Economic

violence

Total

Percentage of respondents who experienced violence over their lifetime

Women 41.0 42.4 30.0 48.0 28.2 71.5

Men 34.4 50.8 23.1 51.9 25.8 71.4

Percentage of respondents who experienced violence in the 12 months prior to the survey

Women 20.8 8.9 10.6 22.5 14.4 42.9

Men 18.7 10.2 9.5 28.0 12.2 43.6


Table 5 shows that 71.5 per cent of women and 71.4 per cent of men reported having

experienced at least one form of violence (domestic and non-domestic) over their lifetime,

and 42.9 per cent of women and 43.6 per cent of men experienced at least one form of

violence in the 12 months prior to the survey. The following subsections discuss in detail

how each of these forms of violence was experienced by the respondents of the GFLHS

2015 across their characteristics and geographical location.

4.1.1. Social violence

Table 5 shows that 41.0 per cent of female respondents and 34.4 per cent of male

respondents experienced at least one type of social violence over their lifetime, committed

by perpetrators both within and outside domestic relations. Just over 20 per cent of women

and 18.7 per cent of men experienced at least one form of social violence during the 12

months prior to the survey. The survey explored further the incidence of social violence

along its different dimensions (as defined above), and across regional and socio-economic

63

characteristics of the sample. This disaggregated information is included in Tables 6–9 and

discussed in the following paragraphs.

Dimensions of social violence. Social violence experienced by both men and women over

their lifetime was mostly in the form of rumours, photos or videos. Almost a quarter (24.0

per cent) of women (Table 6) and 22.8 per cent of men (Table 8) experienced this form of

violence. The second and third most common forms of social violence over the lifetime

were, respectively, not being allowed outside their homes (experienced by 14.4 per cent of

women and 10.6 per cent of men), and being kept away from family and friends

(experienced by 14.1 per cent of women and 8.9 per cent of men). These patterns were

similar when considering the incidence of social violence in the 12 months prior to the

survey (Tables 7 and 9).

Age patterns. The incidence of social violence among men and women was strongly

associated with age, with older people being less likely to have experienced social violence.

For example, 53 per cent and 49.4 per cent of women aged 15–19 and 20–24 years,

respectively, reported having experienced social violence over their lifetime. This was in

contrast with 33.1 per cent of women aged 40–49 years and 33.7 per cent of women aged

50–60 years (Table 6). There were also large differences by age among men: almost 40 per

cent of young men aged 15–19 years and 36.6 per cent of men aged 20–24 years reported

having experienced social violence over their lifetime (Table 8). The incidence of social

violence over the lifetime of men aged 40–49 years was 29.7 per cent. It was 32.2 per cent

among men aged 50–60 years. These age patterns were similar when considering the

incidence of social violence in the 12 months prior to the survey (Tables 7 and 9).

Employment status. The incidence levels of social violence over the lifetime of women and

men were slightly higher for women not working (i.e. not in waged employment): 42.8 per

cent of currently non-working women reported having experienced social violence,

compared to 40.5 per cent of self-employed women and 40.3 per cent of employed women

(Table 6). Social violence was most prominent among employed men (37.6 per cent),

followed by men who were not working at the time of the survey (34.6 per cent) and self-

employed men (32.7 per cent) (Table 8). Tables 7 and 9 show similar patterns across

employment status of women and men for the incidence of social violence in the 12 months

prior to the survey.

Marital status. The results indicate that married or divorced, separated or widowed men

and women were more likely than those never married to have experienced social violence.

Almost 41 per cent of women who were married or living with a partner, and 41.5 per cent

of women divorced, separated or widowed, experienced social violence, in comparison to

39.9 per cent of women who were never married (Table 6). Thirty per cent of never-married

men reported having experienced at least one form of social violence over their lifetime, in

comparison to, respectively, 35.1 and 35.8 per cent of men who lived with a partner (or

were married) or men who were divorced, separated or widowed (Table 8). Women who

were never married (32.8 per cent) reported higher levels of social violence in the 12

64

months prior to the survey than women with other marital status (this was the reverse of

the pattern observed for lifetime social violence) (Table 7).

Geographical patterns. The incidence of lifetime social violence was higher among women

living in rural areas (42.3 per cent) than women living in urban areas (39.9 per cent). It was

also higher among women living in the Eastern (67.7 per cent) and Western regions (48.3

per cent) (Table 6) than for those living in other regions. The incidence of lifetime social

violence among men showed a similar pattern (Table 8). Differences across urban and rural

areas were less pronounced among women for the 12 months prior to the survey than

among women reporting lifetime experiences of social violence. Among the former,

women living in urban areas were slightly more likely than women in rural areas to have

experienced this form of violence (21.5 per cent, in comparison to 20.1 per cent in rural

areas) (Table 7). Men living in urban areas were also less likely to report having

experienced social violence over the 12 months prior to the survey than men in rural areas

(21.5 per cent, in comparison to 16.1 per cent of men living in rural areas – a reverse of the

pattern observed among men (and women) reporting social violence over their lifetime)

(Table 9). These differences may be due to recent changes in media technology (since the

main form of social violence was the spread of rumours, videos and photos), which are

more likely to be more widespread in urban areas.

Education levels and asset ownership. The incidence of lifetime social violence was lower

among women with no formal education (31.8 per cent) and among women with higher

(university) education (35.1 per cent) than among women with intermediate levels of

education (primary, middle, secondary and technical) (Table 6). Men with no education or

with technical education were less likely to report social violence (25.6 per cent and 26.2

per cent, respectively), than men with other levels of education (Table 8). The results show

no clear association between the lifetime incidence of social violence and asset ownership

for men or women. These education and wealth patterns were similar when considering

the incidence of social violence in the 12 months prior to the survey (Tables 7 and 9).

65

Table 6: Percentage of women who have experienced domestic or non-domestic social violence over their lifetime

Kept from

seeing

friends or

family

Stopped from

leaving home

Asked to

report

activities

Stalked Rumour,

photos or

videos spread

Forced to

have an

abortion

Any social

violence

Age group (years)

15–19 23.0 27.8 15.4 4.2 24.6 6.7 53.0 20–24 20.1 20.7 18.1 6.0 24.9 9.4 49.4 25–29 16.7 13.2 12.7 4.2 27.6 12.7 47.7 30–39 11.1 12.4 10.0 2.9 24.1 8.5 39.0 40–49 10.8 10.3 5.2 1.2 20.7 7.2 33.1 50–60 11.6 13.3 5.2 1.8 22.1 5.1 33.7

Employment

Self-employed 13.7 14.0 8.6 2.7 25.2 8.2 40.5 Employed 11.4 11.3 13.8 3.5 21.5 9.5 40.3 Not working 16.2 16.8 13.4 4.2 21.8 9.1 42.8

Marital status

Never married 18.8 19.9 10.1 1.6 17.6 1.5 39.9 Married or living together 13.4 14.0 10.7 2.9 23.9 8.8 40.9 Divorced/separated/widowed 15.2 14.2 9.1 4.2 26.1 9.3 41.5

Residence

Urban 13.2 12.4 11.6 3.8 22.9 8.9 39.9 Rural 15.0 16.6 8.8 2.4 25.2 8.2 42.3

66

Kept from

seeing

friends or

family

Stopped from

leaving home

Asked to

report

activities

Stalked Rumour,

photos or

videos spread

Forced to

have an

abortion

Any social

violence

Region

Western 14.9 16.0 6.0 1.1 32.3 12.7 48.3 Central 12.4 15.4 7.7 2.4 18.6 11.5 38.1 Greater Accra 8.9 6.7 9.6 3.0 17.0 6.1 29.4 Volta 10.7 10.9 9.3 3.6 37.3 4.9 47.3 Eastern 35.2 41.0 17.1 6.3 42.0 11.9 67.7 Ashanti 7.0 9.0 12.1 2.4 25.9 11.7 42.2 Brong Ahafo 17.2 10.1 13.8 4.4 16.2 9.7 40.3 Northern 19.5 19.7 4.9 2.9 13.2 0.9 31.0 Upper East 15.1 19.2 3.3 2.7 14.7 0.0 27.3 Upper West 8.0 7.5 12.6 1.1 11.1 2.3 24.6

Education level None 13.0 12.9 5.1 1.5 18.7 5.3 31.8

Primary 17.5 18.5 10.1 3.5 29.3 9.8 46.2 Middle/JSS/JHS 14.1 14.9 10.6 3.3 26.7 10.7 44.5 Secondary 13.0 12.8 16.8 4.3 19.9 7.6 42.6 Technical 12.7 11.3 11.9 3.3 28.8 13.5 43.8 Higher 11.6 10.3 14.1 5.0 18.2 3.6 35.1

Asset quintile

Lowest 15.4 16.2 6.9 2.5 24.8 6.9 41.3 Second 15.0 15.7 9.7 3.5 24.4 8.8 41.2 Middle 12.3 13.5 10.6 3.3 26.1 11.6 43.8 Fourth 14.0 13.5 12.2 3.5 23.0 9.0 39.4 Highest 12.9 11.9 12.7 2.8 21.4 6.3 39.6

Total 14.1 14.4 10.3 3.1 24.0 8.5 41.0


67

Table 7: Percentage of women who experienced domestic or non-domestic social violence in the last 12 months

Kept from

seeing

friends or

family

Stopped from

leaving home

Asked to

report

activities

Stalked Rumour,

photos or

videos spread

Forced to

have an

abortion

Any social

violence

Age group (years)

15–19 16.9 21.8 12.2 3.9 17.9 3.9 44.2 20–24 7.8 7.1 12.7 3.7 16.7 2.1 31.5 25–29 5.0 4.3 8.4 2.4 17.7 0.9 25.7 30–39 2.4 2.3 4.2 1.7 15.2 0.4 18.7 40–49 1.3 1.5 2.3 0.8 9.6 0.0 12.7 50–60 0.9 0.7 0.7 0.4 8.2 0.0 9.8

Employment


Marital status


Residence

Urban 4.1 3.6 6.6 1.9 14.4 0.7 21.5 Rural 4.0 4.7 4.7 1.8 13.5 1.0 20.1

68

Kept from

seeing

friends or

family

Stopped from

leaving home

Asked to

report

activities

Stalked Rumour,

photos or

videos spread

Forced to

have an

abortion

Any social

violence

Region


Education level

None 2.4 2.2 1.7 1.3 8.5 0.3 12.5 Primary 4.1 4.9 4.9 1.5 17.9 0.7 23.6 Middle/JSS/JHS 5.0 5.3 5.7 2.0 16.3 1.3 23.7 Secondary 4.5 4.6 13.1 3.4 13.9 0.5 26.9 Technical 0.0 1.3 7.5 1.7 11.3 0.6 15.8 Higher 4.8 2.4 7.4 1.2 10.4 0.8 17.0

Asset quintile


Total 4.1 4.1 5.7 1.9 14.0 0.8 20.8


69

Table 8: Percentage of men who have experienced domestic or non-domestic social violence over their lifetime

Kept from

seeing

friends or

family

Stopped from

leaving home

Asked to

report

activities

Stalked Rumour,

photos or

videos spread

Forced to

have an

abortion

Any social

violence

Age group (years)

15–19 11.8 19.0 5.5 2.5 18.5 N/A 39.8 20–24 11.4 10.0 15.0 4.5 24.1 N/A 36.6 25–29 8.6 10.1 14.0 4.4 22.4 N/A 36.3 30–39 6.8 9.2 10.5 2.5 24.6 N/A 33.9 40–49 7.9 8.0 4.3 2.9 22.5 N/A 29.7 50–60 9.6 11.1 4.3 2.0 22.4 N/A 32.2

Employment

Self-employed 9.9 11.6 7.8 2.1 22.1 N/A 32.7 Employed 7.1 8.2 13.1 5.2 26.2 N/A 37.6 Not working 8.8 11.2 8.5 3.1 20.4 N/A 34.6

Marital status

Never married 10.6 13.6 5.4 2.1 15.5 N/A 30.0 Married or living together 9.0 10.5 9.7 2.8 24.0 N/A 35.1 Divorced/separated/widowed 7.0 8.1 12.2 5.6 25.1 N/A 35.8

Residence

Urban 7.7 8.4 10.0 3.6 22.5 N/A 33.8 Rural 10.1 12.6 8.8 2.8 23.1 N/A 35.0

70

Kept from

seeing

friends or

family

Stopped from

leaving home

Asked to

report

activities

Stalked Rumour,

photos or

videos spread

Forced to

have an

abortion

Any social

violence

Region

Western 5.9 9.3 10.7 6.3 34.8 N/A 49.3 Central 10.1 9.6 9.0 4.1 25.1 N/A 35.0 Greater Accra 5.3 3.6 9.6 2.6 12.0 N/A 22.9 Volta 2.2 5.7 7.6 1.7 33.5 N/A 39.6 Eastern 28.4 33.1 10.4 4.4 34.7 N/A 59.2 Ashanti 4.1 7.9 11.8 3.9 24.3 N/A 33.9 Brong Ahafo 11.1 11.0 12.8 3.1 21.4 N/A 33.0 Northern 13.2 14.3 4.9 0.6 11.9 N/A 21.5 Upper East 12.2 11.8 4.0 1.0 6.1 N/A 17.4 Upper West 6.1 6.1 1.8 0.0 14.9 N/A 19.2

Education level

None 10.6 14.5 3.0 0.5 14.1 N/A 25.6 Primary 12.7 14.4 5.8 1.3 25.4 N/A 37.0 Middle/JSS/JHS 8.1 11.1 7.7 2.5 23.2 N/A 34.6 Secondary 8.4 9.3 13.4 6.6 29.6 N/A 41.0 Technical 7.8 1.2 9.0 4.0 17.9 N/A 26.2 Higher 7.1 6.3 17.4 4.1 19.6 N/A 32.8

Asset quintile

Lowest 12.0 13.9 5.7 1.9 20.2 N/A 31.5 Second 9.8 11.9 7.5 2.5 22.3 N/A 34.5 Middle 7.3 8.9 8.6 2.2 26.9 N/A 37.1 Fourth 7.6 10.0 12.0 4.0 24.3 N/A 35.6 Highest 7.0 6.9 14.6 5.7 21.0 N/A 33.7

Total 8.9 10.6 9.4 3.2 22.8 N/A 34.4


71

Table 9: Percentage of men who experienced domestic or non-domestic social violence in the last 12 months

Kept from

seeing

friends or

family

Stopped from

leaving home

Asked to

report

activities

Stalked Rumour,

photos or

videos spread

Forced to

have an

abortion

Any social

violence

Age group (years)

15–19 5.3 7.1 4.1 1.7 13.0 N/A 23.1 20–24 2.3 1.0 10.6 3.5 16.4 N/A 26.1 25–29 1.0 1.4 8.1 3.5 12.1 N/A 20.5 30–39 0.6 1.5 5.7 1.5 13.3 N/A 18.2 40–49 0.8 0.7 2.1 1.8 11.2 N/A 13.9 50–60 0.8 1.8 2.2 2.0 8.5 N/A 12.1

Employment


Marital status


Residence

Urban 1.7 2.3 6.4 2.8 13.6 N/A 21.5 Rural 1.3 1.5 4.9 1.9 11.4 N/A 16.1

72

Kept from

seeing

friends or

family

Stopped from leaving home

Asked to

report

activities

Stalked Rumour,

photos or videos spread

Forced to

have an

abortion

Any social

violence

Region


Education level


Asset quintile


Total 1.5 1.9 5.6 2.3 12.5 N/A 18.7


73

4.1.2. Physical violence

Men reported higher incidence levels of physical violence than women (Table 5): 42.4 per

cent of female respondents and 50.8 per cent of male respondents experienced at least one

of the above acts of physical violence over their lifetime. Almost 9 per cent of women and

10.2 per cent of men experienced at least one act of physical violence during the 12 months

prior to the survey. Tables 10–13 show further disaggregated detail on the incidence of

physical violence across different population groups and characteristics, which are

discussed in the paragraphs below.

Dimensions of physical violence. The most common forms of physical violence

experienced by both men and women over their lifetime were being slapped or being hit

by thrown objects. This form of violence affected 32.7 per cent of women (Table 10) and

38.3 per cent of men (Table 12). The next most prominent form of physical violence was

being hit by another person, experienced by 15.1 per cent of women (Table 10) and 20.8 per

cent of men (Table 12). These patterns of physical violence were very similar across the

lifetime of individuals and in the 12 months prior to the survey.

Age patterns. The incidence of physical violence over people’s lifetimes was strongly

associated with age for both women and men, with older individuals less likely to have

experienced physical violence. Over 47 per cent of women aged 24 years and below

reported having experienced physical violence, compared to 40.6 per cent of women aged

40–49 years and 35.7 per cent of women aged 50–60 years (Table 10). Younger men aged

15–19 years were particularly at risk of experiencing physical violence: 64.3 per cent of men

in this age group reported having experienced at least one form of physical violence over

their lifetime, compared to, for instance, 42.9 per cent of men aged 50–60 years (Table 12).

These age patterns were very similar across the lifetime of individuals and in the 12 months

prior to the survey (Tables 11 and 13).

Employment status. Employed and non-working women were more likely to have

experienced physical violence over their lifetime (43.7 and 43.6 per cent, respectively), than

self-employed women (41.8 per cent) (Table 10). This pattern was similar among the male

sample, but incidence levels were higher: 56.9 per cent of men who were not working

reported having experienced at least one form of physical violence over their lifetime,

while 55 per cent of employed and 46 per cent of self-employed men reported experiencing

this (Table 12). The employment patterns were similar to those observed when considering

experiences of physical violence in the 12 months prior to the survey (Table 11 and 13).

Marital status. The lifetime incidence of physical violence was higher among individuals,

particularly women, who were divorced, separated or widowed, or married. Experiences

with at least one form of physical violence were reported by 46.3 per cent of divorced,

separated or widowed women and by 41.8 per cent of women who were married (or living

with a partner), in contrast to 35.1 per cent of women who were never married (Table 10).

The differences in the results for the male sample were less pronounced (Table 12): 52.6 per

74

cent of divorced, separated or widowed men reported having experienced physical

violence, compared to 51.4 per cent of men who were never married and 50.1 per cent who

were married. The incidence of physical violence in the 12 months prior to the survey was

highest among never-married women (Table 11).

Geographical patterns. The incidence of lifetime physical violence was higher among

women living in rural areas (44.4 per cent) than women living in urban areas (40.7 per cent)

(Table 10). A similar proportion of men in rural and urban areas experienced physical

violence (50.7 per cent and 50.9 per cent, respectively) (Table 12). Women and men in the

Eastern region reported having experienced higher levels of physical violence than those

in any other region. These patterns were common across the lifetime of individuals and in

the 12 months prior to the survey (Table 11 and 13).

Education levels and asset ownership. Women and men with no formal education (34.6

and 31.7 per cent, respectively) and women and men with higher education (38.0 and 51.2

per cent, respectively) were less likely to experience physical violence than women and

men with primary education (45.3 and 52.1 per cent, respectively) or other levels of

education (Tables 10 and 12). This suggests that intermediate levels of education may be

associated with increased risk of physical violence among women and men. Women in the

highest and fourth asset quintiles were less likely to experience physical violence (38.9 and

41.4 per cent, respectively) than women in the three lowest asset quintiles. For instance,

44.7 per cent of women in the lowest asset quintile reported experiencing physical violence

(Table 10). The opposite was true for men: men in the lowest asset quintile were less likely

to experience physical violence (46.1 per cent) than men in the highest asset quintile (50.1

per cent) (Table 12). These education and wealth patterns were very similar across the

lifetime of individuals and in the 12 months prior to the survey (Tables 11 and 13).

75

Table 10: Percentage of women who have experienced domestic or non-domestic physical violence over their lifetime

Slapped or

thrown

things at

Pushed

or shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any physical

violence

Age group (years)

15–19 35.0 18.5 13.5 16.8 2.4 1.7 1.3 47.6 20–24 36.2 15.0 14.7 15.2 1.7 1.2 2.2 47.0 25–29 34.7 18.9 14.5 14.2 2.7 1.7 2.6 45.5 30–39 32.0 13.3 15.3 13.7 2.1 0.7 4.1 42.0 40–49 32.1 14.6 17.0 12.7 2.3 1.0 2.8 40.6 50–60 28.5 12.3 14.2 11.5 2.1 0.5 1.7 35.7

Employment

Self-employed 32.8 14.5 16.1 13.6 2.6 1.1 2.8 41.8 Employed 29.6 17.5 14.4 13.6 1.8 0.3 3.9 43.7 Not working 33.9 15.1 12.6 14.0 1.4 1.0 2.3 43.6

Marital status

Never married 22.3 17.9 10.4 9.4 0.0 1.3 1.3 35.1 Married or living together 32.2 14.1 15.2 13.7 2.1 0.9 2.8 41.8 Divorced/separated/widowed 36.9 17.2 15.5 14.5 3.3 1.2 3.3 46.3

Residence

Urban 31.6 14.0 15.4 11.6 1.8 0.8 2.9 40.7 Rural 34.0 16.0 14.7 16.0 2.7 1.2 2.7 44.4

76

Slapped or

thrown

things at

Pushed

or shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any physical

violence

Region

Western 35.5 8.8 5.6 9.0 0.4 0.0 2.3 40.5 Central 36.0 11.4 16.0 13.0 5.2 0.5 3.7 44.2 Greater Accra 21.1 12.1 13.1 5.1 1.1 0.9 2.9 29.8 Volta 44.4 31.0 29.0 22.7 2.5 1.0 0.8 56.3 Eastern 42.8 25.3 29.4 28.2 4.9 3.9 4.6 61.9 Ashanti 37.6 11.9 11.0 16.5 1.3 0.4 3.3 46.3 Brong Ahafo 35.5 16.1 15.9 10.2 3.0 1.5 3.1 44.2 Northern 25.9 15.8 12.4 17.0 0.5 0.4 1.3 39.0 Upper East 11.2 5.5 3.2 4.4 2.4 1.1 0.0 13.5 Upper West 13.9 4.3 6.1 7.4 3.1 0.0 2.4 19.9

Education level

None 27.0 14.5 14.6 12.7 2.5 0.5 1.8 34.6

Primary 35.5 15.9 14.6 17.0 3.3 1.8 4.0 45.3 Middle/JSS/JHS 37.3 15.5 16.9 13.8 2.3 1.2 2.9 47.0 Secondary 28.2 12.5 13.2 13.0 1.1 1.2 1.6 40.3 Technical 34.8 19.5 17.5 10.3 0.0 0.0 5.9 45.2 Higher 25.5 13.5 9.9 10.8 0.9 0.0 3.6 38.0

Asset quintile

Lowest 36.1 20.2 18.7 15.6 3.9 0.9 2.3 44.7 Second 32.7 14.6 11.9 15.4 2.7 1.1 3.1 44.0 Middle 33.6 13.8 14.1 13.1 1.5 0.9 2.3 42.9 Fourth 29.0 12.5 17.4 12.5 1.3 1.0 2.5 38.9 Highest 33.0 13.9 13.7 10.6 1.6 1.2 4.0 41.4

Total 32.7 15.0 15.1 13.7 2.2 1.0 2.8 42.4


77

Table 11: Percentage of women who experienced domestic or non-domestic physical violence in the last 12 months

Slapped or

thrown

things at

Pushed

or shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Age group (years)

15–19 17.7 7.8 6.1 7.8 1.7 1.1 1.1 28.0 20–24 10.1 3.1 4.1 2.5 0.2 0.3 0.0 14.1 25–29 6.5 3.9 2.2 2.6 1.4 0.3 1.0 9.5 30–39 6.2 1.2 1.7 2.5 0.6 0.0 1.6 9.2 40–49 3.2 0.8 1.1 0.5 0.2 0.0 0.2 3.8 50–60 1.0 0.0 0.2 0.0 0.0 0.0 0.0 1.2

Employment


Marital status


Residence

Urban 6.0 1.7 2.3 2.1 0.5 0.1 0.9 8.6 Rural 6.4 2.6 1.7 2.2 0.7 0.2 0.6 9.3

78

Slapped or

thrown

things at

Pushed or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Region


Education level

None 3.5 1.3 1.1 0.6 0.6 0.0 0.3 4.3


Asset quintile


Total 6.2 2.1 2.1 2.1 0.6 0.2 0.7 8.9


79

Table 12: Percentage of men who have experienced domestic or non-domestic physical violence over their lifetime

Slapped or

thrown

things at

Pushed or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Age group (years)

15–19 47.2 30.9 22.8 15.9 3.7 2.1 3.7 64.3 20–24 38.6 17.9 20.1 16.9 1.4 3.1 5.1 54.2 25–29 37.4 16.5 16.9 12.0 3.6 2.1 6.9 47.5 30–39 37.9 18.3 21.6 14.5 4.6 1.6 10.2 50.9 40–49 39.0 20.8 22.6 16.7 3.3 1.3 8.3 49.5 50–60 32.6 22.6 21.9 17.5 3.3 1.9 3.8 42.9

Employment


Marital status


Residence

Urban 38.3 19.0 19.8 13.7 3.9 2.0 8.0 50.9 Rural 38.2 21.4 21.8 16.9 2.9 1.9 5.9 50.7

80

Slapped or

thrown

things at

Pushed or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Region


Education level

None 23.8 15.4 15.6 12.3 2.8 1.4 3.2 31.7 Primary 39.7 23.3 20.7 13.2 2.2 1.6 6.3 52.1 Middle/JSS/JHS 39.4 20.0 20.8 15.6 4.1 2.3 7.5 52.1 Secondary 44.4 21.3 25.0 16.0 3.7 2.3 6.7 59.2 Technical 44.4 19.7 19.9 14.8 0.0 1.4 6.8 57.8 Higher 38.3 21.2 20.1 18.8 3.7 1.5 9.8 51.2

Asset quintile


Total 38.3 20.2 20.8 15.3 3.4 2.0 6.9 50.8


81

Table 13: Percentage of men who experienced domestic or non-domestic physical violence in the last 12 months

Slapped or

thrown

things at

Pushed or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Age group (years)

15–19 18.6 14.9 7.8 3.3 3.1 0.8 2.0 30.6 20–24 7.2 3.0 3.0 2.1 0.8 0.5 1.0 14.2 25–29 4.9 2.0 1.1 0.4 1.5 0.0 0.4 8.7 30–39 3.9 2.4 1.6 0.8 0.3 0.0 1.7 7.3 40–49 2.1 1.8 2.2 0.5 0.5 0.0 0.5 5.1 50–60 1.0 1.4 1.2 0.3 0.3 0.8 0.7 4.0

Employment


Marital status


Residence

Urban 5.0 3.8 2.2 0.6 0.7 0.6 0.8 9.7 Rural 5.7 3.1 2.6 1.5 1.1 0.0 1.2 10.6

82

Slapped

or thrown

things at

Pushed or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Region


Education level


Asset quintile


Total 5.4 3.5 2.4 1.0 0.9 0.3 1.0 10.2


83

4.1.3. Sexual violence

Table 5 shows that 30 per cent of women and 23.1 per cent of men experienced sexual

violence at least once over their lifetime, and 10.6 per cent of women and 9.5 per cent of

men reported having experienced sexual violence at least once over the 12 months that

preceded the survey. Tables 14 to 17 provide further disaggregation of the data on the

incidence of sexual violence, as discussed below.

Dimensions of sexual violence. The most prevalent form of sexual violence among both

women and men over their lifetimes was being subject to sexual comments (18 per cent

among women and 12.1 per cent among men) (Tables 14 and 16). This was followed by

unwanted touches (15.3 per cent among women and 10.3 per cent among men) and being

physically forced to have sex (9 per cent among women and 6.9 per cent among men).

These patterns of sexual violence were very similar across the lifetime of individuals and

in the 12 months prior to the survey (Tables 15 and 17).

Age patterns. Similar to social and physical violence, the incidence of lifetime sexual

violence was higher among younger women: 38.2 per cent of women aged 15–19 years,

40.4 per cent of women aged 20–24 years and 38.3 per cent of women aged 30–39 years

reported having experienced at least one act of sexual violence. In contrast, 18.8 per cent of

women aged 50–60 years experienced these forms of violence (Table 14). The incidence of

sexual violence was highest among men aged 30–39 years (29.7 per cent) and men aged 20–

24 years (29.2 per cent). The lowest incidence was again among the oldest male age group

(13.8 per cent) (Table 16). These age patterns were also observed when considering the 12

months prior to the survey (Tables 15 and 17).

Employment status. Employed women and men (41.9 and 31.5 per cent, respectively) were

more likely to have experienced sexual violence than men and women not in employment

(Tables 14 and 16). This is an issue that will be explored further when focusing the analysis

on domestic violence. These patterns were very similar across the lifetime of individuals

and in the 12 months prior to the survey (Tables 15 and 17).

Marital status. As observed with social and physical violence, the incidence of sexual

violence was more prominent among women and men who were divorced, separated or

widowed at the time of the survey (31.7 per cent of women and 29.4 per cent of men) (Table

14). The differences in relation to other marital status were considerable, with 22.6 per cent

of never-married women and 16.9 per cent of never-married men having reported sexual

violence (Table 16). As with other forms of violence, the incidence of sexual violence in the

12 months prior to the survey was highest among never-married women (Table 17). This

may indicate a generational effect, which will be explored further in the next sections.

Geographical patterns. Contrary to other forms of violence analysed above, the incidence

of sexual violence was consistently higher among women and men living in urban areas

(32.1 per cent of women and 28.5 per cent of men, compared with 27.8 and 18.1 per cent in

84

rural areas, respectively) (Tables 14 and 16). As with other forms of violence, the incidence

of sexual violence among women was higher in the Eastern region (45.5 per cent) than in

any other region. The highest levels of incidence of sexual violence among men were

observed in the Ashanti region (32.4 per cent). The lowest incidence of sexual violence

against women and men was in the Upper East region (8.8 per cent and 6.7 per cent,

respectively). These geographical patterns were very similar when considering the

incidence of sexual violence in the 12 months prior to the survey (Tables 15 and 17).

Education levels and asset ownership. Women with no formal education were about half

as likely as women with some level of education to report sexual violence. Women with

secondary (39.1 per cent) and higher education (37.1 per cent) were more likely to have

experienced sexual violence than women with primary (32.8 per cent) or middle levels of

education (32.0 per cent) (Table 14). Men with higher levels of education were more likely

to have experienced sexual violence over their lifetime: 34.9 per cent, compared to 20.1 per

cent among men with primary education. Men with no formal education were the least

likely to report sexual violence over their lifetime (7.8 per cent) (Table 16). Reflecting this

latter pattern, sexual violence against men was also more prominent at the top end of the

asset ownership distribution. The proportion of men experiencing any type of sexual

violence over their lifetime was twice as high for male respondents in the highest asset

quintile (29.3 per cent) than for those in the lowest quintile (15.1 per cent) (Table 16). This

association was similar for women, but the differences across quintiles were less

pronounced (Table 14). These patterns were also observed when considering the incidence

of sexual violence in the 12 months prior to the survey (Tables 15 and 17).

85

Table 14: Percentage of women who have experienced domestic or non-domestic sexual violence over their lifetime

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Any

sexual

violence

Age group (years)

15–19 20.6 26.4 10.3 3.5 3.3 2.0 0.9 38.2 20–24 25.6 18.6 11.8 1.4 7.9 3.4 5.1 40.4 25–29 19.8 20.2 14.0 2.1 6.6 4.1 6.1 38.3 30–39 17.8 14.2 8.9 1.6 5.9 2.7 4.6 29.5 40–49 13.8 10.5 6.2 1.4 3.5 0.5 4.1 21.4 50–60 13.8 9.6 3.2 1.6 2.1 1.3 2.5 18.8

Employment


Marital status


Residence

Urban 19.0 16.0 10.1 1.6 5.3 2.9 5.2 32.1 Rural 16.9 14.5 7.8 1.9 4.9 1.8 3.2 27.8

86

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Any

sexual

violence

Region


Education level


Asset quintile


Total 18.0 15.3 9.0 1.7 5.1 2.4 4.3 30.0


87

Table 15: Percentage of women who experienced domestic or non-domestic sexual violence in the last 12 months

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Any

sexual

violence

Age group (years)

15–19 11.9 15.4 3.1 0.9 0.4 0.5 1.6 22.1 20–24 13.1 8.1 2.2 0.1 1.6 1.8 1.1 20.2 25–29 7.3 6.1 1.8 0.5 0.9 1.8 1.4 13.7 30–39 7.0 2.5 0.8 0.2 0.7 1.1 0.8 10.1 40–49 2.3 2.0 0.2 0.0 0.1 0.4 0.0 4.2 50–60 1.9 0.9 0.0 0.3 0.0 0.0 0.0 2.1

Employment


Marital status


Residence

Urban 7.6 4.3 1.1 0.1 0.5 1.5 1.0 11.7 Rural 5.4 4.5 1.1 0.4 0.7 0.5 0.4 9.3

88

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Any

sexual

violence

Region


Education level

None 1.8 1.6 0.5 0.4 0.1 0.0 0.1 3.6


Asset quintile


Total 6.6 4.4 1.1 0.3 0.6 1.0 0.7 10.6


89

Table 16: Percentage of men who have experienced domestic or non-domestic sexual violence over their lifetime

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Any

sexual

violence

Age group (years)

15–19 14.0 5.8 2.6 0.8 0.8 0.0 0.7 19.4 20–24 17.3 15.5 9.9 2.4 2.6 0.5 3.7 29.2 25–29 11.4 10.1 7.0 3.4 3.3 2.0 3.8 25.1 30–39 14.6 14.0 8.0 2.7 2.8 1.4 3.7 29.7 40–49 7.0 7.6 6.4 0.5 1.6 2.2 1.3 16.5 50–60 8.0 5.0 5.6 1.2 1.4 0.2 0.8 13.8

Employment


Marital status


Residence

Urban 16.1 12.8 8.2 3.0 2.9 1.1 2.4 28.5 Rural 8.3 7.8 5.7 1.0 1.7 1.3 2.7 18.1

90

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Any

sexual

violence

Region


Education level


Asset quintile


Total 12.1 10.3 6.9 2.0 2.3 1.2 2.6 23.1


91

Table 17: Percentage of men who experienced domestic or non-domestic sexual violence in the last 12 months

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Any

sexual

violence

Age group (years)

15–19 10.8 4.8 1.9 0.8 0.8 0.7 0.0 14.5 20–24 7.9 6.3 1.8 1.6 0.3 1.6 0.5 14.2 25–29 5.8 4.7 1.7 1.8 0.8 1.1 0.4 12.2 30–39 6.0 3.3 0.7 0.9 0.3 1.1 0.7 10.2 40–49 2.6 1.4 2.1 0.0 0.0 0.0 0.2 5.8 50–60 0.0 0.0 0.7 0.0 0.0 0.0 0.0 0.7

Employment


Marital status


Residence

Urban 7.8 4.3 1.7 1.4 0.5 0.7 0.3 12.8 Rural 2.9 2.5 1.1 0.4 0.2 0.9 0.4 6.4

92

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Any

sexual

violence

Region


Education level


Asset quintile


Total 5.3 3.4 1.4 0.9 0.4 0.8 0.3 9.5


93

4.1.4. Psychological violence

Psychological violence was the most common type of violence experienced by women and

men across Ghana. About half of the respondents (51.9 per cent of men and 48 per cent of

women) experienced an act of psychological violence at least once in their lifetime (Table

5), while 22.5 per cent of women and 28 per cent of men experienced at least one type of

psychological violence in the 12 months prior to the survey (Table 6). Further

disaggregated information on the incidence of psychological violence is provided in Tables

18–21 and discussed in the paragraphs below.

Dimensions of psychological violence. The most common type of psychological violence

reported by both men and women over their lifetime was being insulted, humiliated or

belittled. Almost 46 per cent of women (Table 18) and 48.8 per cent of men (Table 20) in the

sample reported having experienced this form of violence. Other forms of psychological

violence affected only small numbers of women and men in Ghana over their lifetime and

in the 12 months prior to the survey (Tables 19–21).

Age patterns. The results show that age was less strongly correlated to lifetime experiences

of psychological violence than to the other types of violence discussed above, with no clear

pattern of association for either women or men (Tables 18 and 20). But, in line with

discussions above for other forms of violence, younger women and men (aged 15–19 years)

were more likely to have reported at least one incident of psychological violence in the 12

months prior to the survey than all other older age groups: 38.8 per cent and 38.9 per cent

of women and men, respectively, aged 15–19 years experienced psychological violence in

that period, in comparison to 14.8 per cent of women and 18.8 per cent of men aged 50–60

years (Tables 19 and 21).

Employment status. The incidence of lifetime experiences of psychological violence was

highest among self-employed women (49.5 per cent) and employed men (56.1 per cent). It

was lowest among women who were not working (44.7 per cent) and among self-employed

men (49.2 per cent) (Tables 18 and 20). These employment patterns were also observed

when considering the incidence of psychological violence in the 12 months prior to the

survey (Tables 19–21).

Marital status. Men and women who were divorced, separated or widowed were about 6

percentage points more likely to have experienced psychological violence over their

lifetime than all other marital statuses on average (Tables 18 and 20). The lifetime incidence

of psychological violence across marital status was similar to what was found for the other

forms of violence analysed so far, and when considering the 12 months prior to the survey

(Tables 19–21).

Geographical patterns. Women and men living in rural areas (50.3 and 52.8 per cent,

respectively) were more likely to have reported at least one type of psychological violence

over their lifetime than those living in urban areas (45.8 and 51 per cent, respectively). As

94

with other forms of violence, the incidence of psychological violence was highest in the

Eastern region (73.7 per cent for women and 70.3 per cent for men). The lowest incidence

of lifetime psychological violence was observed in the Upper East region for women (13.1

per cent) and in the Upper West region for men (23.9 per cent) (Tables 18 and 20). These

geographical patterns were similar when considering the incidence of psychological

violence in the 12 months prior to the survey (Tables 19–21).

Education levels and asset ownership. Women without formal education and women with

higher education were up to 6.8 percentage points less likely to report having experienced

psychological violence than women with intermediate levels of education (primary,

middle/junior secondary school (JSS)/junior high school (JHS), secondary and technical

education) (Table 18). The incidence of psychological violence among men steadily

increased with the level of education until secondary/technical education. Men with higher

levels of education were as likely as those with primary education to experience

psychological violence (Table 20). Women in the highest asset quintile were 7.3 percentage

points less likely to report having experienced psychological violence over their lifetime

than women in the lowest asset quintile, whereas the incidence of psychological violence

was highest among men in the highest asset quintile (54.5 per cent). These patterns were

similar when considering the 12 months prior to the survey (Tables 19–21).

95

Table 18: Percentage of women who have experienced domestic or non-domestic psychological violence over their lifetime

Insulted,

humiliated or

belittled

Ignored or

threatened to be

abandoned

Scared or

intimidated

on purpose

Threatened

with the use

of a weapon

Threatened to

be hurt

Any

psychological

violence

Age group (years)

15–19 45.9 11.3 3.3 1.8 0.3 49.3 20–24 46.0 10.1 4.3 1.0 0.8 47.9 25–29 48.2 8.5 4.7 2.1 0.8 50.1 30–39 42.4 7.8 4.3 3.6 1.8 45.0 40–49 48.4 8.5 6.6 2.6 1.1 50.5 50–60 47.0 7.0 7.6 1.9 2.4 47.6

Employment

Self-employed 47.5 8.4 5.5 2.6 1.7 49.5 Employed 42.6 9.8 6.6 3.6 0.6 46.2 Not working 43.0 7.9 3.8 1.4 0.9 44.7

Marital status

Never married 39.9 4.1 2.6 0.8 0.0 41.0 Married or living together 44.7 7.7 5.0 2.1 1.0 46.7 Divorced/separated/widowed 51.4 12.1 6.7 4.2 3.1 54.0

Residence

Urban 44.0 8.4 5.3 2.1 1.0 45.8 Rural 48.1 8.4 5.1 2.9 1.8 50.3

96

Insulted,

humiliated or

belittled

Ignored or

threatened to be

abandoned

Scared or

intimidated

on purpose

Threatened

with the use

of a weapon

Threatened to

be hurt

Any

psychological

violence

Region

Western 54.0 4.7 1.7 1.1 1.0 56.7 Central 45.6 16.3 3.7 2.1 1.8 50.2 Greater Accra 36.4 7.2 5.3 2.3 1.0 37.7 Volta 59.1 10.9 20.7 1.2 2.9 59.5 Eastern 71.1 12.3 7.0 5.6 3.0 73.7 Ashanti 47.6 8.2 1.1 2.6 0.9 48.9 Brong Ahafo 38.1 8.5 5.2 3.7 0.7 41.4 Northern 41.7 2.8 5.2 1.0 1.0 42.8 Upper East 12.0 0.0 1.1 0.0 1.1 13.1 Upper West 14.9 5.3 1.5 1.2 0.0 17.3

Education level

None 42.4 7.5 6.0 2.3 1.4 44.1 Primary 49.7 9.5 4.3 2.9 2.7 51.5 Middle/JSS/JHS 47.5 9.1 4.9 2.4 0.8 49.8 Secondary 43.5 7.5 4.3 1.0 0.8 45.4 Technical 49.5 6.1 10.5 8.2 4.0 49.5 Higher 42.0 7.8 7.1 2.4 1.1 45.1

Asset quintile

Lowest 51.8 10.2 8.8 3.2 2.4 53.5 Second 44.2 9.0 3.6 2.9 1.3 46.8 Middle 50.9 7.1 3.6 2.1 0.6 52.6 Fourth 40.7 6.3 3.8 1.0 1.1 42.3 Highest 43.7 9.8 7.4 3.2 1.6 46.2

Total 45.9 8.4 5.2 2.4 1.4 48.0


97

Table 19: Percentage of women who experienced domestic or non-domestic psychological violence in the last 12 months

Insulted,

humiliated

or belittled

Ignored or

threatened to be

abandoned

Scared or

intimidated

on purpose

Threatened

with the use

of a weapon

Threatened to

be hurt

Any

psychological

violence

Age group (years)

15–19 37.0 7.7 1.9 1.0 0.0 38.8 20–24 24.1 5.1 0.7 0.0 0.4 26.4 25–29 24.4 3.0 1.4 1.0 0.3 25.6 30–39 19.8 3.0 0.9 1.2 0.8 21.5 40–49 16.6 2.3 1.2 0.6 0.4 19.0 50–60 14.2 0.7 0.8 0.6 0.0 14.8

Employment


Marital status


Residence

Urban 21.1 3.2 1.2 0.6 0.3 22.4 Rural 20.7 2.9 0.8 1.0 0.6 22.6

98

Insulted,

humiliated

or belittled

Ignored or

threatened to be

abandoned

Scared or

intimidated

on purpose

Threatened

with the use

of a weapon

Threatened to

be hurt

Any

psychological

violence

Region


Education level


Asset quintile


Total 20.9 3.1 1.1 0.8 0.4 22.5


99

Table 20: Percentage of men who have experienced domestic or non-domestic psychological violence over their lifetime

Insulted,

humiliated

or belittled

Ignored or

threatened to be

abandoned

Scared or

intimidated

on purpose

Threatened

with the use

of a weapon

Threatened to

be hurt

Any

psychological

violence

Age group (years)

15–19 50.2 6.8 8.2 2.0 0.8 55.6 20–24 50.6 7.5 6.1 3.2 2.5 53.7 25–29 47.4 4.8 5.9 1.7 3.2 51.0 30–39 45.9 6.8 9.7 6.9 2.7 49.6 40–49 48.3 6.2 6.9 3.2 1.9 49.5 50–60 53.0 7.6 9.1 2.6 2.5 55.4

Employment


Marital status


Residence

Urban 47.5 6.5 7.4 3.8 3.1 51.0 Rural 50.0 6.6 8.0 3.4 1.8 52.8

100

Insulted,

humiliated

or belittled

Ignored or

threatened to be

abandoned

Scared or

intimidated

on purpose

Threatened

with the use of

a weapon

Threatened

to be hurt

Any

psychological

violence

Region


Education level


Asset quintile


Total 48.8 6.5 7.7 3.6 2.4 51.9


101

Table 21: Percentage of men who experienced domestic or non-domestic psychological violence in the last 12 months

Insulted,

humiliated

or belittled

Ignored or

threatened to be

abandoned

Scared or

intimidated

on purpose

Threatened

with the use

of a weapon

Threatened to

be hurt

Any

psychological

violence

Age group (years)

15–19 36.1 4.0 2.4 1.2 0.8 38.9 20–24 34.0 4.3 2.0 1.5 1.5 38.2 25–29 28.0 1.5 2.5 0.5 1.7 29.5 30–39 22.4 2.9 3.2 1.7 1.1 25.2 40–49 20.9 2.0 0.8 0.6 0.5 22.5 50–60 17.9 1.8 0.6 0.3 0.0 18.8

Employment


Marital status


Residence

Urban 26.1 2.4 1.8 1.2 1.5 28.7 Rural 25.4 2.9 2.2 0.8 0.5 27.4

102

Insulted,

humiliated

or belittled

Ignored or

threatened to be

abandoned

Scared or

intimidated

on purpose

Threatened

with the use

of a weapon

Threatened to

be hurt

Any

psychological

violence

Region


Education level


Asset quintile


Total 25.7 2.7 2.0 1.0 1.0 28.0


103

4.1.5. Economic violence

About one quarter of respondents – 28.2 and 25.8 per cent of women and men, respectively

– experienced an act of economic violence at least once during their lifetime; about one in

eight men and women experienced at least one type of economic violence in the 12 months

preceding the survey (Table 5). Tables 22–25 provide disaggregated information on the

incidence of economic violence across different population groups and socio-economic

characteristics, which is discussed in the paragraphs below.

Dimensions of economic violence. The most common form of economic violence

experienced by both women and men over their lifetime was the denial of chop money: 20.2

per cent of women and 12.2 per cent of men reported this form of violence (Tables 22 and

24). This was also the most prevalent form of violence reported by women and men in the

12 months prior to the survey (11.6 and 5.7 per cent, respectively) (Tables 23 and 25).

Age patterns. Younger age is again a risk factor for lifetime economic violence, particularly

among men. Around 30 per cent of men aged 15–29 years reported having experienced

economic violence in their lifetime, compared to 23.9 per cent of men aged 30–39 years,

21.6 per cent of men aged 40–49 years and 20.1 per cent of men aged 50–60 years (Table

24). The association between age and incidence of economic violence was less pronounced

among women. The highest incidence was among women aged 25–29 years (33.9 per cent),

while the lowest was among women aged 50–60 years (23.6 per cent) (Table 22). These age

patterns were also observed when considering the incidence of economic violence in the

12 months prior to the survey (Table 23–25).

Employment status. Just over 30 per cent of women who were not working at the time of

the survey experienced economic violence, compared to 28.3 per cent of self-employed

women and 23.1 per cent of employed women (Table 22). In contrast, economic violence

was more prevalent among employed men: 29.4 per cent of employed men reported having

experienced economic violence, compared to 28.3 per cent of men not working and 22.9

per cent of self-employed men (Table 24). These results suggest that employed men were

at higher risk of economic violence, perhaps because economic expectations may be higher

among these men. On the other hand, economic violence was more prevalent among

women who stay at home, perhaps as a result of their economic dependence on men, as

discussed in the literature surveyed in Section 2. The same patterns were observed when

the survey questions referred to the 12 months prior to the survey (Tables 23–25).

Marital status. As with other forms of violence, lifetime experiences of economic violence

were also more prominent among men and women who were divorced, separated or

widowed at the time of the survey. Thirty-three per cent of divorced, separated or

widowed women experienced economic violence, compared to 18.6 per cent of never-

married women and 27.6 per cent of married women or women living with a partner (Table

22). Almost 35 per cent of divorced, separated or widowed men experienced economic

violence, compared to 25.2 per cent of never-married men and 23.4 per cent of men married

104

or living with a partner (Table 24). These patterns were similar when considering the 12

months prior to the survey (Tables 23–25).

Geographical patterns. The incidence of economic violence was higher among women

living in rural areas (31.3 per cent) than women living in urban areas (25.5 per cent) (Table

22), whereas there was almost no difference in the lifetime incidence of economic violence

among men living in urban or rural areas (Table 24). As with other forms of violence, both

women and men living in the Eastern region (48.7 per cent and 51.9 per cent, respectively)

were more likely than individuals in other regions to have experienced at least one form

of economic violence over their lifetime. This distribution of the incidence of economic

violence across locations was similar for the 12 months prior to the survey (Tables 23–25).

Education levels and asset ownership. The incidence of economic violence was lower

among women with higher education (19.4 per cent) than among women with no formal

education or intermediate levels of education (primary, middle/JSS/JHS, secondary,

technical). Women with primary education were more likely than women with other

education levels to have experienced economic violence (35.8 per cent) (Table 22). In

contrast, the incidence of economic violence was lowest among men without formal

education (10.9 per cent). Wealthier women in the two highest asset quintiles were around

5–10 percentage points less likely to have experienced economic violence over the course

of their lives. In contrast, men in the four highest asset quintiles were more likely to

experience economic violence (26.7, 27.3, 27.1 and 26 per cent, respectively) than men in

the lowest asset quintile (22 per cent) (Table 24). These education and wealth patterns were

similar when considering the 12 months prior to the survey (Tables 23–25).

105

Table 22: Percentage of women who have experienced domestic or non-domestic economic violence over their lifetime

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Age group (years)

15–19 19.4 1.4 3.4 1.6 2.5 6.8 1.1 30.4 20–24 21.3 2.5 1.9 3.6 2.4 6.2 0.1 31.7 25–29 25.9 3.6 2.6 5.2 3.9 4.7 0.0 33.9 30–39 20.4 2.6 2.1 2.6 3.0 3.9 0.2 27.4 40–49 15.8 2.2 1.8 3.2 2.7 2.6 0.1 24.2 50–60 16.8 2.3 1.8 3.6 1.9 3.6 0.1 23.6

Employment


Marital status


Residence

Urban 16.9 2.0 2.6 4.1 2.7 4.3 0.2 25.5 Rural 23.8 3.2 1.7 2.7 3.1 4.3 0.2 31.3

106

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Region


Education level


Asset quintile


Total 20.2 2.6 2.2 3.4 2.8 4.3 0.2 28.2


107

Table 23: Percentage of women who experienced domestic or non-domestic economic violence in the last 12 months

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced to

work

Denied

food

Any

economic

violence

Age group (years)

15–19 15.0 1.4 1.6 0.9 0.4 2.8 5.5 20.4 20–24 13.8 1.1 1.0 1.9 1.1 2.2 1.9 17.9 25–29 17.1 1.5 1.0 2.0 1.9 2.0 1.3 20.8 30–39 12.6 1.2 1.1 0.9 1.4 0.3 1.3 15.4 40–49 7.9 0.5 0.3 0.0 0.9 0.5 0.2 8.9 50–60 4.2 0.3 0.1 0.3 0.5 0.0 0.3 5.6

Employment


Marital status


Residence

Urban 9.3 0.6 1.0 1.2 1.0 0.7 0.7 12.1 Rural 14.2 1.5 0.5 0.7 1.5 1.4 1.9 17.0

108

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Region


Education level


Asset quintile


Total 11.6 1.0 0.8 1.0 1.2 1.0 1.3 14.4


109

Table 24: Percentage of men who have experienced domestic or non-domestic economic violence over their lifetime

Denied

pocket

money

Cash

taken out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced to

work

Denied

food

Any

economic

violence

Age group (years)

15–19 21.4 4.7 4.1 4.0 1.3 6.6 0.1 32.3 20–24 17.2 4.2 0.6 6.4 4.5 7.4 0.1 29.2 25–29 13.4 4.5 0.4 5.8 3.1 9.6 0.1 30.0 30–39 8.7 3.8 1.4 9.1 3.8 6.2 0.3 23.9 40–49 9.2 3.7 2.2 7.2 3.7 2.9 0.7 21.6 50–60 8.2 4.2 1.8 6.2 0.8 4.2 0.4 20.1

Employment


Marital status


Residence

Urban 12.4 3.8 2.1 7.8 2.9 6.9 0.3 25.9 Rural 12.0 4.4 1.1 5.8 3.2 5.6 0.3 25.7

110

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Region


Education level


Asset quintile


Total 12.2 4.1 1.6 6.8 3.1 6.2 0.3 25.8


111

Table 25: Percentage of men who experienced domestic or non-domestic economic violence in the last 12 months

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced to

work

Denied

food

Any

economic

violence

Age group (years)

15–19 15.1 3.4 3.5 3.1 0.5 3.8 2.9 23.3 20–24 9.5 3.0 0.1 3.8 3.1 2.7 3.2 18.9 25–29 5.5 1.5 0.0 2.2 1.4 5.0 0.9 14.2 30–39 2.2 0.8 0.4 3.2 1.2 1.1 0.3 8.3 40–49 3.9 0.5 0.6 2.2 0.9 0.5 1.0 8.5 50–60 3.4 0.9 0.8 1.2 0.3 0.7 0.6 5.9

Employment


Marital status


Residence

Urban 5.5 1.4 1.2 3.1 1.2 2.9 1.2 13.3 Rural 6.0 1.6 0.2 2.1 1.3 1.5 1.4 11.3

112

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Region


Education level


Asset quintile


Total 5.7 1.5 0.7 2.6 1.3 2.2 1.3 12.3


113

4.2. Incidence of domestic violence against women and men in Ghana

This subsection and the remainder of the report will focus on violence experienced or

perpetrated within domestic relations – domestic violence – in the 12 months prior to the

survey.24 Domestic violence is defined, as per the Ghana Domestic Violence Act 732, as any

form of violence against men or women perpetrated by a domestic relation. Domestic

relations are core members of the family, as well as non-family members who live within

the household. Table 26 summarises the overall incidence of domestic violence across the

five types of violence in the 12 months prior to the survey.

The most common form of domestic violence reported by women in the 12 months prior

to the GFLHS 2015 was economic violence (12.8 per cent), followed by social violence (11.6

per cent) and psychological violence (9.3 per cent) (Table 26). This order is different from

the previous section, where psychological violence had the highest incidence (Table 5).

Domestic violence among men was mainly psychological violence (7.9 per cent), followed

by social violence (7.7 per cent) and economic violence (7.3 per cent) (Table 26). This order

is the same as for the overall levels of violence discussed in the previous section (Table 5).

Table 26: Incidence of domestic violence in Ghana 2015 in the 12 months prior to the

survey (in percentages)

Social

violence

Physical

violence

Sexual

violence

Psychological

violence

Economic

violence

Total

(at least

one type)

Women 11.6 6.0 2.5 9.3 12.8 27.7

Men 7.7 2.1 1.4 7.9 7.3 20.0


It is important to note that these types of domestic violence did not occur in isolation.

According to the GFLHS 2015, many respondents who reported having experienced

domestic violence experienced multiple forms and types of violence: 23.3 per cent of

women and 18.9 per cent of men who experienced domestic violence reported having

experienced two types of domestic violence, while 9.4 per cent of women and 5.7 per cent

of men reported having experienced three different types of domestic violence. In what

follows, the analysis focuses on each type of violence for ease of discussion.

24 Qualitative and quantitative data on domestic violence covered only the 12 months prior to the survey. No

questions were asked about lifetime experiences due to budget and time constraints, as this would have

increased substantially the length of the questionnaires and interviews.

114

4.2.1. Domestic social violence

Table 27 compares the proportion of domestic and non-domestic social violence. The

results show that almost all types of social violence against men and women were largely

perpetrated within domestic relations. The only exception was the spread of rumours,

photos and videos without permission: only 23.4 per cent and 13.2 per cent of this form of

violence against women and men, respectively, constituted domestic violence. Nearly all

(97.4 per cent) of the instances reported by women of being prevented from seeing friends

or family were perpetrated by domestic relations. The following forms of violence against

women were committed mostly by domestic relations: not being allowed to leave their

home (98.5 per cent), required to report on their activities (88.4 per cent), being stalked

(62.5 per cent) and being forced to have an abortion (women only) (84.6 per cent). The

results for men were similar, with the exception of stalking, which was mostly (59 per cent)

perpetrated by non-domestic relations.

Table 27: Percentage of domestic social violence over the last 12 months

Not

seeing

friends

or

family

Not

allowed

to leave

home

Required

to

report

activities

Stalked Rumour,

photos,

videos

spread

Forced to

have an

abortion

Women Perpetrated by family members 63.1 65.4 48.8 31.2 2.4 11.5

Perpetrated by non-family

members but domestic

34.3

33.1

40.9

32.8

21.2

73.1

% Domestic violence 97.4 98.5 88.4 62.5 23.4 84.6

% Non-domestic violence 2.6 1.5 11.6 37.5 76.6 15.4

Total 100.0 100.0 100.0 100.0 100.0 100.0

Men Perpetrated by family members 65.3 80.0 29.8 17.9 3.3 N/A



27.1

15.0

43.3

23.1

10.3

N/A

% Domestic violence 92.4 95.0 73.1 41.0 13.2 N/A

% Non-domestic violence 7.6 5.0 26.9 59.0 86.8 N/A

Total 100.0 100.0 100.0 100.0 100.0 N/A


Tables 28 and 29 illustrate how domestic social violence was distributed across the socio-

economic characteristics and geographical location of women and men, respectively. The

incidence of domestic social violence was higher among women (11.6 per cent) than among

men (7.7 per cent), and all individual acts of domestic social violence were more prominent

among women than among men. The relative importance of each type of domestic social

violence was similar among women and men. The most common form of domestic social

violence against women and men was being required to report their activities (5.1 per cent

115

for women and 3.9 per cent for men), followed by being stopped from leaving home (4.1

per cent for women and 1.6 per cent for men), being kept from seeing friends and family

(4.0 per cent for women and 1.4 per cent for men), being the victim of rumour, photo or

video spread (3.4 per cent for women and 1.7 per cent for men), and being stalked (1.2 per

cent for women and 1.0 per cent for men). In addition, 0.7 per cent of women in the GFLHS

2015 were forced to having an abortion.

Age patterns. As observed with overall levels of social violence, the incidence of domestic

social violence was highest among women in the youngest age group (15–19 years): 35.7

per cent of women in this age group experienced at least one type of social violence in the

12 months prior to the survey. The incidence in other age groups was considerably lower:

19.3 per cent for women aged 20–24 years, 15.3 per cent for women aged 25–29 years, 8.6

per cent for women aged 30–39 years, 5.0 per cent for women aged 40–49 years and 3.9 per

cent for women aged 50–60 years. The incidence of all acts of domestic social violence was

highest among women aged 15–19 years, with the exception of being stalked, which was

highest (2.4 per cent) for women aged 20–24 years. These age patterns were similar for

men, but differences across the age groups were less pronounced: 13.5 per cent of all men

aged 15–19 years experienced at least one form of domestic social violence in the 12 months

prior to the survey, in comparison to 13 per cent of men aged 20–24 years, 8.7 per cent of

men aged 25–29 years, 5.7 per cent of men aged 30–39 years, 3.8 per cent of men aged 40–

49 and 4.4 per cent of men aged 50–60 years.

Employment status. The incidence of domestic social violence was highest for women and

men who were not working at the time of the survey (19.8 per cent for women and 10.9 per

cent for men). Both women and men who were self-employed were at lower risk of

experiencing any act of domestic social violence (8.3 and 5.8 per cent, respectively). All

types of domestic social violence were more commonly reported by women and men who

were not working. The only exceptions were men who were required to report their

activities and men who were victims of a rumour, photo or video being spread. These acts

of domestic social violence were highest among employed men (5.6 per cent and 2 per cent,

respectively).

Marital status. Women who had never been married were at the highest risk of

experiencing any form of domestic social violence: 27.1 per cent of these women

experienced at least one act of domestic social violence in the 12 months prior to the survey,

in comparison with 11 per cent of married women and 9.9 per cent of women who were

divorced, separated or widowed. However, being a victim of the spreading of a rumour,

photo or video or being forced to having an abortion were more common among women

who were divorced, separated or widowed (4.6 per cent and 0.9 per cent, respectively). The

patterns for men were different. Men who were divorced, separated or widowed were

more at risk of experiencing any form of domestic social violence (8.3 per cent, in

comparison to 8.1 per cent for never-married men and 7.4 per cent for married men).

However, never-married men were more likely to report being stopped from leaving home

(3.8 per cent) and being kept from seeing friends and family (2.4 per cent). These latter

116

results are likely to be explained by the fact that these men lived at home with their parents,

an issue that will be explored in the next section, which focuses on the perpetration of

violence.

Geographical patterns. There was little difference in the levels of domestic social violence

experienced by women living in urban and rural areas (11.9 and 11.3 per cent,

respectively). The differences were more pronounced for men: 8.5 per cent of men living

in urban areas experienced at least one act of domestic social violence, in comparison to 6.9

per cent of men living in rural areas. Women and men living in the Western region were

more likely to report having experienced at least one act of domestic social violence (14.8

and 12.2 per cent, respectively). The incidence of domestic social violence was lowest

among women living in the Brong Ahafo region (7.4 per cent) and men living in the Upper

East region (1.7 per cent).

Education levels and asset ownership. Table 29 shows a linear association between

domestic social violence and the education and wealth status of men: the higher their level

of education and asset quintile, the higher the likelihood of men reporting having

experienced at least one act of domestic social violence in the 12 months prior to the survey.

Thus 10.4 per cent of men with higher education reported having experienced domestic

social violence, in comparison to 2.6 per cent of men with no education. Similarly, 11.9 per

cent of men in the highest asset quintile experienced domestic social violence, in

comparison to 4.0 per cent of men in the lowest asset quintile. As discussed before, this

result may be explained by the fact that better educated and wealthier men may also be

more likely to report experiences of domestic violence. The patterns for women were less

pronounced. Women with secondary levels of education (15.8 per cent) and in the middle

asset quintile (12.9 per cent) were the most likely to report having experienced domestic

social violence. The incidence of domestic social violence was lowest among women with

no education (5.4 per cent) and women in the highest asset quintile (10.3 per cent).

117

Table 28: Percentage of women who experienced domestic social violence over the last 12 months

Not

seeing

friends or

family

Not

allowed to

leave

home

Required

to

report

activities

Stalked Rumour,

photos,

videos

spread

Forced to

have an

abortion

Any

domestic

social

violence

Age group (years)

15–19 16.5 21.8 12.2 2.1 5.8 1.9 35.7 20–24 7.6 7.1 9.4 2.4 3.6 1.9 19.3 25–29 4.8 4.1 7.8 1.5 4.7 0.9 15.3 30–39 2.4 2.3 4.0 1.1 3.5 0.4 8.6 40–49 1.3 1.5 2.0 0.6 2.0 0.0 5.0 50–60 0.9 0.7 0.7 0.4 2.1 0.0 3.9

Employment


Marital status


Residence

Urban 4.0 3.5 5.7 1.1 3.4 0.7 11.9 Rural 4.0 4.7 4.3 1.3 3.4 0.7 11.3

118

Not

seeing

friends or

family

Not

allowed to

leave

home

Required

to

report

activities

Stalked Rumour,

photos,

videos

spread

Forced to

have an

abortion

Any

domestic

social

violence

Region


Education level

None 2.3 2.2 1.6 0.6 1.6 0.2 5.4 Primary 4.0 4.9 4.9 1.5 6.0 0.4 13.4 Middle/JSS/JHS 5.0 5.3 5.1 1.5 4.3 1.1 13.7 Secondary 4.2 4.2 10.9 1.8 2.0 0.3 15.8 Technical 0.0 1.3 4.6 0.0 0.9 0.6 6.8 Higher 4.8 2.4 6.8 0.6 1.5 0.8 10.2

Asset quintile


Total 4.0 4.1 5.1 1.2 3.4 0.7 11.6


119

Table 29: Percentage of men who experienced domestic social violence over the last 12 months

Not

seeing

friends or

family

Not

allowed to

leave

home

Required

to

report

activities

Stalked Rumour,

photos,

videos

spread

Forced to

have an

abortion

Any

domestic

social

violence

Age group (years)

15–19 4.6 6.8 3.3 0.0 3.4 N/A 13.5 20–24 2.2 1.0 6.3 1.7 2.9 N/A 13.0 25–29 1.0 1.4 6.0 2.2 1.3 N/A 8.7 30–39 0.5 0.5 4.0 0.7 1.4 N/A 5.7 40–49 0.8 0.7 1.5 0.7 1.1 N/A 3.8 50–60 0.8 1.8 1.7 0.0 1.1 N/A 4.4

Employment


Marital status


Residence

Urban 1.4 1.8 4.9 1.0 1.3 N/A 8.5 Rural 1.3 1.5 3.0 0.9 2.2 N/A 6.9

120

Not

seeing

friends or

family

Not

allowed to

leave

home

Required

to

report

activities

Stalked Rumour,

photos,

videos

spread

Forced to

have an

abortion

Any

domestic

social

violence

Region


Education level


Asset quintile


Total 1.4 1.6 3.9 1.0 1.7 N/A 7.7


121

4.2.2. Domestic physical violence

All types of physical violence against women were mainly perpetrated by domestic

relations (within and outside the immediate family), whereas all types of physical violence

against men were mainly perpetrated by non-domestic relations (Table 30). Forms of

physical violence against women and men were more likely to be perpetrated by domestic

relations outside the immediate family. The exceptions were women being slapped or

being threatened with a weapon, which were equally likely to be perpetrated by family or

non-family members (36.8 and 38.9 per cent, respectively). It was also more likely that men

were burnt by family members (33.3 per cent) than by non-family members (none).

Table 30: Percentage of domestic physical violence over the last 12 months

Slapped or

thrown

things at

Pushed

or

shoved

Hit Kicked,

dragged

or beaten

up

Choked

or

strangled

Burnt Weapon

used

against

Women

Perpetrated by family

members

36.8 24.2 26.9 35.9 25.0 50.0 38.9

Perpetrated by non-

family members but

domestic

36.8

38.7

31.3

48.4

55.0

50.0

38.9

% Domestic violence 73.0 62.9 58.2 84.4 80.0 100.0 77.8

% Non-domestic

violence

27.0 37.1 41.8 15.6 20.0 0.0 22.2

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Men


members

5.1 0.0 2.0 5.9 0.0 33.3 4.5

Perpetrated by non-

family members but

domestic

22.2

6.4

12.0

23.5

10.5

0.0

18.2

% Domestic violence 27.3 6.4 14.0 29.4 10.5 33.3 22.7 % Non-domestic

violence

72.7 93.6 86.0 70.6 89.5 66.7 77.3

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0


Tables 31 and 32 illustrate the variation of incidence of domestic physical violence across

the socio-economic and geographical characteristics of women and men, respectively.

Women were almost three times more likely to experience physical violence (6 per cent)

than men (2.1 per cent). The most common act of domestic physical violence experienced

by women and men was being slapped or having objects thrown at them (4.4 and 1.4 per

cent, respectively). The second most common form of domestic physical violence for

women was being kicked, dragged or beaten up (1.8 per cent), while for men it was being

hit (0.3 per cent). These results suggest that women and men experienced domestic

122

physical violence in different ways, an issue that will be further analysed in subsequent

sections.

Age patterns. As with other forms of domestic and non-domestic violence, the incidence of

domestic physical violence was highest among younger women (15–19 years old): 15.2 per

cent of these women reported having experienced at least one act of domestic physical

violence, in comparison to 9.4 per cent of women aged 20–24 years, 6.9 per cent of women

aged 25–29 years, 7.1 per cent of women aged 30–39 years, 2.4 per cent of women aged 40–

49 years and 0.5 per cent of women aged 50–60 years. The incidence of all acts of domestic

physical violence was also highest among younger women, with the exception of being

pushed or shoved (highest among women aged 25–29 years: 2.4 per cent) and being hit

(highest among women aged 20–24 years: 2.2 per cent). The association between age and

domestic physical violence was slightly different for men. The incidence of this form of

domestic violence was also highest (8.5 per cent) among younger men (15–19 years).

However, the lowest incidence of domestic physical violence was observed among men

aged 30–39 years (0.5 per cent). In contrast to the data discussed above for women, older

men aged 50–60 years were the most likely to report being hit (0.6 per cent) and being

choked or strangled (0.3 per cent).

Employment status. The incidence of domestic physical violence was highest among

women and men who were not working (9.1 and 3.8 per cent, respectively). The lowest

incidence was observed among women who were employed (4.7 per cent) but among men

who were self-employed (1.5 per cent).

Marital status. Women and men who had never been married at the time of the survey

were the most likely group to report having experienced at least one act of domestic

physical violence in the 12 months prior to the survey (9.0 and 5.6 per cent, respectively).

There were not substantial differences in the number of women and men who were

married or were divorced, separated or widowed and reported having experienced

domestic physical violence. However, women who were divorced, separated or widowed

were more likely than married or never-married women to experience acts of violence:

being pushed or shoved (2.2 per cent), being hit (1.5 per cent), being kicked, dragged or

beaten up (2.9 per cent) or being chocked or strangled (0.9 per cent). These results suggest

that these forms of violence were more likely to be perpetrated within intimate

partnerships. These patterns were observed for women only.

Geographical patterns. Both women and men living in rural areas were more likely to have

experienced domestic physical violence (6.4 and 2.6 per cent, respectively), than those in

urban areas (5.7 and 1.6 per cent, respectively). The highest incidence of domestic physical

violence among women was in the Eastern region (8.7 per cent, in comparison with the

lowest level of 0.8 per cent in the Upper East region). The highest incidence levels among

men were observed in the Central region (4.4 per cent), and the lowest in the Northern

region (0.7 per cent).

123

Education levels and asset ownership. The incidence of domestic physical violence was

highest among women with primary levels of education: 9.2 per cent of these women

experienced domestic physical violence, in comparison to, for instance, 2.9 per cent of

women with no education (who reported the lowest incidence of domestic physical

violence). Women in the middle asset quintile (8.2 per cent) reported higher levels of

domestic physical violence than women in the highest quintile (5.8 per cent) and women

in the lowest quintile (6.3 per cent). These results seem to suggest that women with

intermediate levels of education and wealth may be at a higher risk of experiencing

domestic physical violence. Men with secondary education were more likely to experience

domestic physical violence (3.6 per cent). The lowest levels of incidence were observed

among men with technical education (no men reported domestic physical violence), men

with no education (0.6 per cent) and men with higher education (0.8 per cent). The pattern

for wealth was clearer: men in the two lowest asset quintiles (3.2 and 3.5 per cent,

respectively) were more than three times more likely to experience domestic physical

violence than men in the fourth quintile (0.8 per cent) or in the top asset quintile (1.0 per

cent).

124

Table 31: Percentage of women who experienced domestic physical violence over the last 12 months

Slapped or

thrown

things at

Pushed

or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Age group (years)

15–19 10.4 1.5 1.2 4.9 1.3 1.1 1.1 15.2 20–24 7.2 1.5 2.2 2.1 0.0 0.3 0.0 9.4 25–29 5.5 2.4 1.5 2.6 1.1 0.3 1.0 6.9 30–39 4.7 1.2 1.4 2.4 0.6 0.0 1.0 7.1 40–49 2.0 0.8 0.8 0.3 0.2 0.0 0.2 2.4 50–60 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.5

Employment


Marital status


Residence

Urban 4.4 1.0 1.3 1.7 0.4 0.1 0.6 5.7 Rural 4.4 1.5 1.0 1.9 0.6 0.2 0.5 6.4

125

Slapped or

thrown

things at

Pushed

or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Region


Education level


Asset quintile


Total 4.4 1.2 1.2 1.8 0.5 0.2 0.6 6.0


126

Table 32: Percentage of men who experienced domestic physical violence over the last 12 months

Slapped or

thrown

things at

Pushed

or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Age group (years)

15–19 6.4 0.7 0.5 0.7 0.0 0.8 0.0 8.5 20–24 1.9 0.0 1.0 0.3 0.0 0.0 0.3 3.4 25–29 1.2 0.0 0.2 0.4 0.2 0.0 0.4 2.0 30–39 0.3 0.0 0.0 0.0 0.0 0.0 0.3 0.5 40–49 0.4 0.3 0.0 0.0 0.0 0.0 0.0 0.6 50–60 0.7 0.6 0.6 0.3 0.3 0.0 0.3 0.9

Employment


Marital status


Residence

Urban 1.0 0.1 0.3 0.0 0.0 0.2 0.2 1.6 Rural 1.8 0.3 0.3 0.4 0.2 0.0 0.3 2.6

127

Slapped or

thrown

things at

Pushed

or

shoved

Hit Kicked,

dragged or

beaten up

Choked or

strangled

Burnt Attacked

with

weapon

Any

physical

violence

Region


Education level


Asset quintile


Total 1.4 0.2 0.3 0.2 0.1 0.1 0.2 2.1


128

4.2.3. Domestic sexual violence

Most types of sexual violence against women were predominantly perpetrated by

domestic relations, with the exception of sexual comments and sexual touches (3.1 and 5.7

per cent of these two acts, respectively, were committed by non-domestic relations). In

contrast, most acts of sexual violence against men – with the exception of sex without

consent and sex without protection – were mainly perpetrated by non-domestic relations.

This finding suggests that sexual violence against women was more prevalent within

domestic relations, whereas for men it was more widespread outside domestic relations

(Table 33). Most domestic sexual violence against women was perpetrated by non-family

members. This includes sexual comments (2.8 per cent perpetrated by non-family

members, and 0.3 per cent by family members), sexual touch (5.7 per cent perpetrated by

non-family members, and none by family members), physically forced to have sex (28.6

per cent perpetrated by non-family members, and 22.9 per cent by family members), and

having sex without protection (69.6 per cent perpetrated by non-family members, and 21.7

per cent perpetrated by family members). All acts of domestic sexual violence against men

were more likely to be committed by non-family members than family members (Table 33).

Table 33: Percentage of domestic sexual violence over the last 12 months

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Sex

because

afraid

Women


members

0.3 0.0 22.9 28.6 37.5 21.7 53.6

Perpetrated by non-

family members but

domestic

2.8

5.7

28.6

14.3

37.5

69.6

25.0


% Non-domestic

violence

96.9 94.3 48.6 57.1 25.0 8.7 21.4

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Men Perpetrated by family

members

1.4 1.4 0.0 0.0 0.0 0.0 6.7

Perpetrated by non-

family members but

domestic

2.2

4.3

12.0

18.8

60.0

75.0

60.0

% Domestic violence 3.6 5.7 12.0 18.8 60.0 75.0 66.7 % Non-domestic

violence

96.4 94.3 88.0 81.2 40.0 25.0 33.3

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0


Tables 34 and 35 illustrate, respectively, how domestic sexual violence was experienced

across different socio-economic groups. As with other forms of domestic violence, women

129

were more likely than men to have experienced domestic sexual violence (2.5 and 1.4 per

cent, respectively). The most common type of domestic sexual violence experienced by

women and men was being forced to have sex because of being afraid of the consequences

of refusing. This form of violence affected 0.8 per cent of all women in the GFLHS 2015 and

0.6 per cent of men. The second most common reason for both women and men was being

forced to have sex without protection (0.7 per cent for women and 0.3 per cent for men).

Age patterns. In contrast to other forms of domestic violence, the incidence of domestic

sexual violence was highest among women aged 20–24 years: 4.9 per cent of women in this

age group reported having experienced domestic sexual violence, in comparison to 4.6 per

cent of women aged 25–29 years, 4.3 per cent of women aged 15–19 years, 2.2 per cent of

women aged 30–39 years and 0.5 per cent of women aged 40–49 years. However, the

incidence of particular types of domestic sexual violence was highest among young

women aged 15–19 years: unwanted sexual touch (0.7 per cent), being physically forced to

have sex (1.5 per cent), and being forced to have sex without protection (1.6 per cent). As

with other forms of domestic violence, the incidence of domestic sexual violence was

lowest among the older women (aged 50–60 years). Men aged 30–39 were at the highest

risk of experiencing domestic sexual violence (2.7 per cent) across almost all acts of sexual

violence reported. The main exception was being the target of an unwanted sexual

comment, which was most prevalent (0.7 per cent) among the younger age group (15–19

years). No men aged 50–60 years reported having experienced an act of sexual violence in

the 12 months prior to the survey.

Employment status. Women who were either employed or not working at the time of the

survey were most likely to report domestic sexual violence (3.0 per cent in both cases).

Employed men were most at risk of domestic sexual violence (also 3.0 per cent). The results

in Tables 34 and 35 show, however, great variation in the association between employment

status and the incidence of domestic violence, with no clear pattern emerging from the

data.

Marital status. The incidence of domestic sexual violence was highest among married

women (2.9 per cent) and divorced, separated or widowed men (2.1 per cent). No never-

married women reported having experienced domestic sexual violence. The incidence of

domestic sexual violence was lowest among never-married men (0.4 per cent). The results

show, however, a variation in how women and men experienced different acts of domestic

sexual violence. Notably, women who were divorced, separated or widowed experienced

the highest incidence of unwanted sexual comments (0.6 per cent) and acts of forced sex

(0.2 per cent). Married men experienced the highest incidence of unwanted sexual touches

(0.2 per cent) and acts of physically forced sex (0.3 per cent), while never-married men

reported the highest incidence of unwanted sexual comments (0.4 per cent).

Geographical patterns. Similar to domestic physical violence, women and men living in

urban areas reported a higher incidence of domestic sexual violence (2.8 and 1.6 per cent,

respectively) than those in rural areas. The highest level of incidence of domestic sexual

130

violence among women and men was observed in the Central region (5.8 and 2.5 per cent,

respectively). The lowest incidence of domestic sexual violence among women was in the

Northern and Upper East regions (both 0.6 per cent), and no men in the Brong Ahafo,

Upper East or Upper West regions reported domestic sexual violence.

Education levels and asset ownership. The incidence of domestic sexual violence was

highest among women with secondary education: 4.9 per cent of these women experienced

at least one act of sexual violence in the 12 months prior to the survey, in comparison to,

for instance, 0.9 per cent of women with no education. Men with technical or higher

education were more likely to have experienced domestic sexual violence (2.2 and 2.1 per

cent, respectively, in comparison with men with no education). Both men and women in

the two highest asset quintiles were more at risk of experiencing domestic sexual violence

(3.9 and 2.5 per cent for women, and 1.7 and 2.7 per cent for men) than men and women in

the other asset quintiles.

131

Table 34: Percentage of women who experienced domestic sexual violence over the last 12 months

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Had sex

because

afraid

Any

sexual

violence

Age group (years)

15–19 0.0 0.7 1.5 0.0 0.4 1.6 0.5 4.3 20–24 0.3 0.3 1.1 0.0 1.2 1.1 1.7 4.9 25–29 0.4 0.6 0.9 0.5 0.5 1.0 1.5 4.6 30–39 0.0 0.0 0.6 0.2 0.7 0.8 0.7 2.2 40–49 0.1 0.0 0.1 0.0 0.1 0.0 0.4 0.5 50–60 0.3 0.3 0.0 0.0 0.0 0.0 0.0 0.3

Employment


Marital status


Residence

Urban 0.2 0.2 0.5 0.0 0.3 0.9 1.2 2.8 Rural 0.2 0.3 0.7 0.3 0.7 0.4 0.4 2.2

132

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Had sex

because

afraid

Any

sexual

violence

Region


Education level


Asset quintile


Total 0.2 0.3 0.6 0.1 0.5 0.7 0.8 2.5


133

Table 35: Percentage of men who experienced domestic sexual violence over the last 12 months

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Had sex

because

afraid

Any

sexual

violence

Age group (years)

15–19 0.7 0.0 0.0 0.0 0.0 0.0 0.0 0.7 20–24 0.0 0.0 1.0 0.0 0.3 0.5 1.2 2.3 25–29 0.3 0.2 0.0 0.4 0.4 0.0 0.5 1.4 30–39 0.3 0.1 0.2 0.5 0.3 0.7 1.1 2.7 40–49 0.0 0.2 0.0 0.0 0.0 0.2 0.0 0.3 50–60 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Employment


Marital status


Residence

Urban 0.1 0.2 0.1 0.4 0.2 0.2 0.6 1.6 Rural 0.2 0.0 0.3 0.0 0.2 0.3 0.5 1.2

134

Sexual

comment

Sexual

touch

Physically

forced to

have sex

Otherwise

forced to

have sex

Sex

without

consent

Sex

without

protection

Had sex

because

afraid

Any

sexual

violence

Region


Education level


Asset quintile


Total 0.2 0.1 0.2 0.2 0.2 0.3 0.6 1.4


135

4.2.4. Domestic psychological violence

Table 36 shows that most psychological violence against women was perpetrated within

domestic relations, with the exception of insults (37.1 per cent of this violence was

perpetrated by domestic relations, and 62.9 per cent by non-domestic relations). Most

psychological violence against men was perpetrated outside domestic relations, with the

exception of acts of being ignored or threatened with abandonment (59.6 per cent of this

violence was perpetrated by domestic relations) (Table 36). All acts of psychological

violence against women and men were more likely to be perpetrated by non-family

members than by family members.

Table 36: Percentage of domestic psychological violence over the last 12 months

Insulted,

humiliated,

belittled

Ignored,

threatened

to be

abandoned

Scared,

intimidated

Attacked

with

weapon

Threaten to

be hurt

Women

Perpetrated by family members 13.4 33.7 12.5 26.1 14.3



24.2

46.9

37.5

52.2

50.0

% Domestic violence 37.1 80.6 50.0 78.3 64.3

% Non-domestic violence 62.9 19.4 50.0 21.7 35.7

Total 100.0 100.0 100.0 100.0 100.0

Men Perpetrated by family members 7.6 17.3 2.8 0.0 5.6



20.7

42.3

8.3

15.0

16.7

% Domestic violence 27.6 59.6 11.1 15.0 22.2

% Non-domestic violence 72.4 40.4 88.9 85.0 77.8

Total 100.0 100.0 100.0 100.0 100.0


Tables 37 and 38 disaggregate the information above further and show how the incidence

of domestic psychological violence was distributed across different socio-economic

characteristics and geographical locations. As with other types of domestic violence,

women were more likely to experience domestic psychological violence (6 per cent) than

men (2.1 per cent). The most common act of domestic psychological violence was being

insulted, humiliated or belittled (7.5 per cent for women and 6.9 per cent for men). The

second most common act was being ignored, or threatened with abandonment (2.4 per cent

for women and 1.5 per cent for men).

Age patterns. The incidence of domestic psychological violence was highest among women

and men in the youngest age group (15–19 years old): 15.2 per cent of women and 8.5 per

cent of men in this age group experienced at least one act of domestic psychological

136

violence in the 12 months prior to the survey. Levels of incidence of domestic psychological

violence reduced as the age of women increased. Among the older age group (50–60 years),

only 0.5 per cent of women reported having experienced domestic psychological violence.

However, the incidence of certain forms of psychological violence, such as being

threatened with a weapon, was highest among women aged 25–29 years (0.9 per cent),

whereas being threatened to be hurt was highest among women aged 30–39 years. The

lowest incidence of domestic psychological violence among men was among those aged

30–39 years (0.5 per cent).

Employment status. Women and men who were not working at the time of the survey

were more likely to experience at least one act of domestic psychological violence (9.1 per

cent for women and 3.8 per cent for men). The lowest incidence of domestic psychological

violence was observed among employed women (4.7 per cent) and self-employed men (1.5

per cent).

Marital status. The incidence of domestic psychological violence was highest among

women and men who were never married (9.0 and 5.6 per cent, respectively). However,

divorced, separated or widowed women were the most likely group to report the incidence

of particular acts of domestic psychological violence, such as being threatened with the use

of a weapon (0.8 per cent) or being threatened to be hurt (0.7 per cent). Likewise, divorced,

separated or widowed men were the most likely to report being insulted, humiliated or

belittled (9.6 per cent) or being threatened to be hurt (0.7 per cent), while married men were

the most likely to report being scared or intimidated on purpose (0.3 per cent) or being

threatened with a weapon (0.3 per cent).

Geographical patterns. Women and men living in rural areas were more likely to have

reported being a victim of domestic psychological violence (6.4 and 2.6 per cent,

respectively) than those in urban areas (5.7 and 1.6 per cent, respectively). The incidence

of domestic psychological violence among women was highest in the Eastern region (8.7

per cent) and lowest in the Upper East region (0.8 per cent). Among men, it was highest

among those living in the central region (4.4 per cent) and lowest among those living in the

Northern region (0.7 per cent).

Education levels and asset ownership. The incidence of domestic psychological violence

was highest among women with primary levels of education (9.2 per cent), and lowest

among women with no education (2.9 per cent) or women with the highest levels of

education (3.8 per cent). Men with secondary education were those most likely to report

experiences of domestic psychological violence (3.6 per cent). As with women, men with

no education (0.6 per cent) or men with the highest levels of education (technical: none,

and higher: 0.8 per cent) were the least likely to report having experienced at least one act

of domestic psychological violence in the 12 months prior to the survey. The incidence of

domestic psychological violence was highest among women in the middle asset quintile

(8.2 per cent) and men in the second asset quintile (3.5 per cent). Women and men in the

fourth and fifth asset quintiles were the least likely to report experiences of domestic

137

psychological violence (3.7 and 5.8 per cent, respectively, for women, and 0.8 and 1.0 per

cent for men). These results suggest that women and men at intermediate levels of

education and wealth were those most at risk of experiencing domestic psychological

violence.

138

Table 37: Percentage of women who experienced domestic psychological violence over the last 12 months

Insulted,

humiliated,

belittled

Ignored,

threatened

to be

abandoned

Scared,

intimidated

Attacked

with

weapon

Threaten

to be hurt

Any

psychological

violence

Age group (years)

15–19 20.0 6.8 1.4 1.0 0.0 15.2 20–24 10.5 4.2 0.3 0.0 0.2 9.4 25–29 8.5 2.3 0.3 0.9 0.3 6.9 30–39 7.0 2.1 0.5 0.7 0.6 7.1 40–49 3.5 1.9 0.2 0.6 0.1 2.4 50–60 4.3 0.7 0.6 0.6 0.0 0.5

Employment


Marital status


Residence

Urban 6.7 2.7 0.4 0.4 0.2 5.7 Rural 8.5 2.2 0.6 0.9 0.3 6.4

139

Insulted,

humiliated,

belittled

Ignored,

threatened

to be

abandoned

Scared,

intimidated

Attacked

with

weapon

Threaten

to be hurt

Any

psychological

violence

Region


Education level


Asset quintile


Total 7.5 2.4 0.5 0.6 0.3 6.0


140

Table 38: Percentage of men who experienced domestic psychological violence over the last 12 months

Insulted,

humiliated,

belittled

Ignored,

threatened

to be

abandoned

Scared,

intimidated

Attacked

with

weapon

Threaten

to be hurt

Any

psychological

violence

Age group (years)

15–19 11.5 3.2 0.7 0.0 0.0 8.5 20–24 10.2 3.3 0.0 0.3 0.4 3.4 25–29 7.8 0.8 0.2 0.0 0.4 2.0 30–39 4.2 1.0 0.0 0.3 0.5 0.5 40–49 6.7 1.1 0.3 0.0 0.0 0.6 50–60 3.8 0.9 0.3 0.3 0.0 0.9

Employment


Marital status


Residence

Urban 6.4 1.5 0.0 0.2 0.4 1.6 Rural 7.4 1.6 0.4 0.2 0.1 2.6

141

Insulted,

humiliated,

belittled

Ignored,

threatened

to be

abandoned

Scared,

intimidated

Attacked

with

weapon

Threaten

to be hurt

Any

psychological

violence

Region


Education level


Asset quintile


Total 6.9 1.5 0.2 0.2 0.2 2.1


142

4.2.5. Domestic economic violence

For most acts of economic violence against women, the majority were perpetrated by

domestic relations, with the exception of cash being taken and property damaged: 34.4 and

46.4 per cent of these acts, respectively, were committed by domestic relations (Table 39).

A large part of economic violence against men was perpetrated outside domestic relations,

with three exceptions: 83.2 per cent of violence in the form of being denied pocket money,

92.3 per cent of violence in the form of having belongings controlled, and 52.2 per cent of

violence in the form of being prohibited from working were perpetrated by domestic

relations (Table 39). All forms of domestic economic violence against men were more likely

to be committed by non-family members than by family members. Most forms of domestic

economic violence against women were also committed by non-family members, with the

exception of having their belongings controlled (60.9 per cent committed by family

members, and 34.8 per cent by non-family members) and being prohibited from working

(48.6 per cent committed by family members, and 28.6 per cent by non-family members).

Table 39: Percentage of domestic economic violence over the last 12 months

Denied

pocket

money

Cash

taken out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced to

work

Denied

food

Women


members

44.1 15.6 60.9 3.6 48.6 30.0 36.4

Perpetrated by non-

family members but

domestic

52.3

18.8

34.8

42.9

28.6

33.3

59.1


% Non-domestic

violence

4.0 65.6 8.7 53.6 25.7 36.7 4.6

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Men


members

28.7 15.1 38.5 5.7 0.0 7.7 23.1

Perpetrated by non-

family members but

domestic

54.5

18.2

53.8

15.1

52.2

25.6

50.0


% Non-domestic

violence

16.8 66.7 7.7 81.1 47.8 66.7 26.9

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0


Tables 40 and 41 illustrate how domestic economic violence was distributed across the

socio-economic characteristics and geographical location of women and men in Ghana.

Domestic economic violence was experienced by 12.8 per cent of women and 7.3 per cent

of men interviewed in the GFLHS 2015. The most common act of domestic economic

143

violence experienced by women and men was being denied pocket money – known in

Ghana as chop money. This act of domestic economic violence was experienced by 11.2 per

cent of women and 4.9 per cent of men. Less than 1 per cent of the sample reported having

experienced any other act of domestic economic violence.

Age patterns. The incidence of domestic economic violence overall was highest among

women aged 25–29 years (19.7 per cent), in comparison to 17.2 per cent among women

aged 15–19 years, 15.9 per cent among women aged 20–24 years, 13 per cent among women

aged 30–39 years, 8.1 per cent among women aged 40–49 years and 5.1 per cent among

women in the oldest age group (50–60 years old). However, women in the youngest age

group (15–19 years) were at a greater risk of particular acts of domestic violence. Notably,

being denied food was highest among this group (5.5 per cent) than among any other age

group (Table 40). These women also reported the highest incidence of being forced to work

(1.5 per cent) and having their belongings controlled (1.6 per cent). The incidence of

domestic economic violence among men was highest among the youngest age group (15–

19 years), at 19 per cent, in comparison to 3.4 per cent among the oldest age group (50–60

years).

Employment status. Women and men who were not working at the time of the survey

were more likely to report having experienced at least one act of domestic economic

violence in the 12 months prior to the survey (16.6 per cent of women and 15.3 per cent of

men). However, the incidence of certain types of domestic violence, such as being

prohibited from working, was highest among self-employed women (1.2 per cent),

whereas employed men were the most likely group to report having their cash taken (0.6

per cent) or being denied food (2.1 per cent). Employed women (8.7 per cent) and self-

employed men (3.9 per cent) were the least likely to report domestic economic violence.

Marital status. The association between marital status and the incidence of domestic

economic violence differs between women and men. Married women (13.4 per cent) were

more likely to experience domestic economic violence than never-married women (10.1

per cent) and women who were divorced, separated or widowed (11.6 per cent). The

incidence of domestic economic violence was highest among never-married men (13.4 per

cent), in comparison to 9.9 per cent among men who were divorced, separated or widowed

and 4.9 per cent among men who were married.

Geographical patterns. The incidence of domestic economic violence was higher among

women living in rural areas (15.9 per cent) than among those living in urban areas (10.4

per cent). The incidence of domestic economic violence was also higher among men living

in rural areas (7.4 per cent), but the difference in relation to men living in urban areas (7.3

per cent) was small. Women living in the Western region (22.8 per cent) and in the Upper

East region (6.7 per cent) were, respectively, the most likely and least likely to report

incidence of domestic economic violence. The incidence of domestic economic violence

was highest among men in the Central region (16.7 per cent) and lowest among men in the

Upper East region (2 per cent).

144

Education levels and asset ownership. Women with primary education were the most

likely to report experiences of domestic economic violence (16.7 per cent). The lowest

incidence of domestic economic violence was observed among women with either higher

education (8.1 per cent) or women with no education (9.6 per cent). Patterns for men were

similar: the incidence of domestic economic violence was lowest among men with no

education (1.4 per cent) or the highest level of education (5.6 per cent), and highest among

men with secondary education (9.6 per cent). Women and men in the second asset quintile

were the most likely to report having experienced domestic economic violence (15.3 and

9.1 per cent, respectively), while women and men in the top asset quintile had the lowest

incidence of domestic economic violence (6.9 and 5.8 per cent, respectively). These results

suggest that individuals with intermediate levels of education and wealth may be at higher

risk of economic violence than better-off and more educated individuals.

145

Table 40: Percentage of women who experienced domestic economic violence over the last 12 months

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Age group (years)

15–19 12.9 0.6 1.6 0.2 0.0 1.5 5.5 17.2 20–24 13.2 0.7 1.0 0.5 1.1 1.2 1.9 15.9 25–29 16.7 0.9 1.0 1.1 1.8 1.3 1.2 19.7 30–39 12.1 0.1 0.8 0.5 1.0 0.1 1.2 13.0 40–49 7.8 0.1 0.3 0.0 0.6 0.5 0.2 8.1 50–60 4.2 0.0 0.1 0.1 0.5 0.0 0.3 5.1

Employment


Marital status


Residence

Urban 8.8 0.3 0.9 0.4 0.6 0.4 0.7 10.4 Rural 13.8 0.5 0.5 0.5 1.4 0.9 1.9 15.6

146

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Region


Education level


Asset quintile


Total 11.2 0.4 0.7 0.4 1.0 0.6 1.2 12.8


147

Table 41: Percentage of men who experienced domestic economic violence over the last 12 months

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Age group (years)

15–19 14.3 1.2 2.9 0.9 0.5 2.4 1.6 19.0 20–24 8.9 1.3 0.1 1.5 1.5 1.8 3.0 13.6 25–29 4.8 0.6 0.0 0.4 0.8 1.2 0.9 7.3 30–39 1.8 0.0 0.4 0.1 0.2 0.0 0.0 2.5 40–49 2.3 0.2 0.6 0.2 0.9 0.3 0.8 4.9 50–60 2.5 0.0 0.8 0.8 0.3 0.0 0.6 3.4

Employment


Marital status


Residence

Urban 4.9 0.2 1.0 0.7 0.5 0.8 1.0 7.3 Rural 4.9 0.8 0.2 0.5 0.9 0.8 1.1 7.4

148

Denied

pocket

money

Cash

taken

out

Belongings

controlled

Property

damaged

Prohibited

from

working

Forced

to work

Denied

food

Any

economic

violence

Region


Education level


Asset quintile


Total 4.9 0.5 0.6 0.6 0.7 0.8 1.0 7.3


149

4.2.6. Summary

The information discussed in the sections above suggests important patterns in how

domestic violence was experienced among men and women, and different socio-economic

groups and geographical locations in Ghana. Some key findings include:

Younger women and men were overall at higher risk than older age groups of

experiencing any form of domestic violence.

Domestic physical or psychological violence appears to have been more prevalent

among non-working, never-married men and women with intermediate levels of

education and wealth and living in rural areas.

Domestic economic violence exhibits a similar pattern as above but was more

prevalent among married women.

Domestic social violence was more prevalent among non-working and urban

women and men with some level of education and wealth.

Domestic sexual violence differs from most other forms of domestic violence: it

appears to be most prominent among employed, better-off, less economically

dependent individuals living in urban areas.

Table 42 summarises these findings, highlighting the categories for which each type of

violence was most prominent. These findings are explored in detail in the remainder of

Section 4.

Table 42: Socio-economic groups with highest incidence of domestic violence

Social

violence

Physical

violence

Sexual

violence

Psychological

violence

Economic

violence

Age (years)

Women 15–19 15–19 20–24 15–19 25–29

Men 15–19 15–19 30–39 15–19 15–19

Employment status

Women Not working Not working Employed;

not working

Not working Not working

Men Not working Not working Employed Not working Not working

Marital status

Women Not married Not married Married Not married Married

Men Divorced Not married Divorced Not married Not married

Residence

Women Urban Rural Urban Rural Rural

Men Urban Rural Urban Rural Rural

Education

Women Secondary Primary Secondary Primary Primary

Men Higher Secondary Technical;

higher

Secondary Secondary

Asset quintile

Women Third Third Fourth Third Second

Men Highest Second Highest Second Second


150

Qualitative data across all regions supported the findings discussed above: both men and

women experienced domestic violence, albeit in different forms and with significant

variations across individual, household, community and regional characteristics. The data

from both the focus group discussions and the in-depth interviews suggested that women

experienced two predominant forms of domestic violence: economic violence (in the form

of men’s refusal to provide chop money) and physical violence (in the form of wife-beating).

The qualitative data showed, in addition and in line with the quantitative results above,

that men experienced domestic violence, by women and sometimes by other men, largely

in more public spaces – streets or marketplaces. The forms of domestic violence that men

experienced in these spaces were predominantly verbal abuse (such as women shouting at

their husbands) and psychological abuse (such as women shaming their husbands in

public). This finding was nationally representative, and was illustrated, for example, by

the following two respondents in the Gbawe mixed FGD (18–25):

Respondent 8: Women in times of violence are physically assaulted, but men are verbally

abused.

Respondent 5: Women are abused physically and verbally, but men are just verbally abused.

4.3. Perpetration of domestic violence in Ghana

The results from the survey, as discussed above, showed that the perpetration of domestic

violence varied considerably between men and women, with women being more likely

than men to have experienced violence perpetrated within domestic relations. This section

discusses in more detail the profiles of domestic violence perpetrators. The section starts

by analysing patterns of violence perpetration within domestic relations (subsection 4.3.1).

Subsection 4.3.2 discusses the profiles of study respondents who reported having

committed social, physical, sexual, psychological and/or economic violence against

another domestic relation. Subsection 4.3.3 focuses on the perpetration of violence against

children and young people.

4.3.1. Violence perpetration within domestic relations

Tables 43 and 44 provide an overview of the main domestic perpetrators of violence for

each type of violence – social, physical, sexual, psychological and economic – as reported

by female and male respondents with respect to the 12 months prior to the survey. It is

important to note that more than one perpetrator could perform different types of violence.

For simplicity, this table shows the aggregate figures of respondents’ first response.

151

Table 43: Perpetrators of domestic violence as reported by women (in percentages)

Panel A: Respondent is head of household (N=1,304)

Social Sexual Physical Psychological Economic

Partner/former partner 48.0 92.0 78.0 34.0 74.0

Spouse 0.0 0.0 5.0 3.0 2.0

In-laws 5.0 3.0 3.0 8.0 2.0

Child 2.0 0.0 0.0 0.0 2.0

Parent 12.0 0.0 1.0 11.0 8.0

Brother, sister 17.0 0.0 5.0 22.0 7.0

Uncle, aunt 2.0 3.0 1.0 6.0 4.0

Cousin, niece, nephew 4.0 0.0 6.0 8.0 0.0

Other relative 9.0 3.0 2.0 9.0 2.0

N 174 36 103 158 183

Panel B: Respondent is a spouse of the head of household (N=958)


Head of household 63.0 96.0 81.0 47.0 83.0


Children 7.0 0.0 1.0 1.0 2.0

Parent of head 11.0 0.0 0.0 17.0 0.0

Parent 3.0 0.0 5.0 0.0 1.0

Brother, sister 4.0 0.0 4.0 5.0 6.0

Uncle, aunt 1.0 0.0 0.0 4.0 2.0


Other relative 8.0 0.0 4.0 17.0 2.0

N 123 24 75 78 122

Panel C: Respondent is the daughter/son of the head of household (N=246)


Parent 65.0 18.0 22.0 51.0 50.0

Brother, sister 9.0 0.0 13.0 16.0 16.0

Partner of head 0.0 0.0 3.0 5.0 0.0


Uncle, aunt 3.0 0.0 0.0 9.0 3.0


Other relative 3.0 18.0 0.0 2.0 3.0

N 118 11 32 43 38


Note: The totals N refer, respectively, to the number of respondents who were the head of the household, the

number of respondents who were spouses of the head of the household and the number of respondents who

were either the daughter or son of the head of the household.

The most predominant perpetrators of domestic violence against women were their

partners or former partners: the share of perpetration by (former) partners ranged between

34 per cent (psychological violence) and 92 per cent (sexual violence) against female

household heads (Table 43, Panel A). The share of husbands perpetrating violence against

their wives ranged between 47 per cent (psychological violence) and 96 per cent (sexual

152

violence) (Table 43, Panel B). Parents, in-laws and brothers and sisters were also identified

as perpetrators of domestic violence. Physical strength was one of the main factors used

by focus group discussion participants across all regions to describe the prevalence of

men’s physical violence against women in the household. In-depth interviews with key

community members confirmed this finding – for example, in the description offered by a

community leader working in the Eastern region, in the East Akim district:

“The difference between how women experience violence and that of men is that the men are

stronger and do harm the women when they beat them.”

Table 44: Perpetrators of domestic violence as reported by men (in percentages)

Panel A: Respondent is head of households (N=1,655)



Spouse 21.0 2.0 1.0 7.0 9.0

In-laws 2.0 0.0 4.0 3.0 0.0

Child 3.0 0.0 2.0 0.0 3.0

Parent 8.0 0.0 18.0 15.0 28.0

Brother, sister 8.0 0.0 18.0 15.0 25.0

Uncle, aunt 2.0 7.0 10.0 10.0 7.0


Other relative 11.0 10.0 18.0 13.0 8.0

N 149 30 49 157 113

Panel B: Respondent is a spouse of the head of the household (N=5)


0.0 0.0 0.0 0.0 0.0

Panel C: Respondent is the daughter/son of the head of household (N=219)


Parent 56.0 0.0 42.0 62.0 84.0

Brother, sister 24.0 0.0 17.0 23.0 13.0


Uncle, aunt 3.0 0.0 8.0 0.0 0.0


Other relative 3.0 0.0 33.0 8.0 0.0

N 34 2 12 26 31


Note: The totals N refer, respectively, to the number of respondents who were the head of the household, the

number of respondents who were spouses of the head of the household and the number of respondents who

were either the daughter or son of the head of the household.

Domestic violence against male respondents was also perpetrated largely by (former)

partners. Economic or physical violence against male heads of household was mainly

perpetrated by parents (28 and 18 per cent, respectively) and siblings (25 and 18 per cent,

respectively) (Table 44). Men who experienced violence committed by parents and siblings

were generally younger men who lived with their family in the same household.

153

4.3.2. Profile of domestic violence perpetrators

The qualitative data from all regions confirmed the range of perpetrators discussed above.

Within the domestic setting, the majority of the perpetrators were men, who committed

violence against their partners (generally women, in heterosexual relationships). As

discussed above, these forms of violence were predominantly economic and physical.

Complementing the quantitative findings shown in Tables 43 and 44, the qualitative

research revealed that domestic violence was both horizontal (between partners) and

vertical (across generations, between caregivers and children). Further, horizontal forms

of violence did not only flow one way, exerted by men against their female partners. The

focus group discussions and in-depth interviews discussed – sometimes at length,

particularly in the Eastern, Ashanti and Volta regions – how men were abused by women,

something that is also revealed in Tables 43 and 44.

Through careful facilitation of the focus group discussions, particularly in groups of

married men and of unmarried men and women, the study found that men also

experienced physical violence. This was not a nationally or regionally representative

finding, but it was a finding reported in urban areas, where men spoke about psychological

and verbal abuse from their female spouses. In the Greater Accra region, one male

participant explained to the rest of the mixed male and female group (18–25 years old) that

women can be stronger than men, and that they can be physically violent to their spouses,

just like men:

“The one I know the woman is much stronger than the man. So she does not obey the rules

of the man. If the man tries to talk she just beats the man up.”

The GFLHS 2015 collected information that allowed the study to explore further the

direction of the perpetration of violence between men and women. Table 45 shows the

share and number of respondents in the survey who reported having perpetrated acts of

domestic violence in the 12 months prior to the survey.

Table 45: Percentage and number of domestic violence perpetrators

Percentage of respondents reporting perpetration of domestic violence

Sex of

perpetrator

Social

violence

Sexual

violence

Physical

violence

Psychological

violence

Economic

violence

Any domestic

violence

Male 0.12 0.06 0.06 0.15 0.06 0.30

Female 0.12 0.01 0.07 0.15 0.03 0.27

Total 0.12 0.03 0.07 0.15 0.04 0.28

154

Number of respondents reporting perpetration of domestic violence

Sex of

perpetrator

Social

violence

Sexual

violence

Physical

violence

Psychological

violence

Economic

violence

Any domestic

violence

Male 222 120 110 277 97 563

Female 362 26 208 422 79 789

Total 584 146 318 699 176 1,352


Table 45 reveals that 563 men (0.3 per cent of the sample) and 789 women (0.27 per cent of

the sample) reported having perpetrated at least one act of domestic violence. Men were

more likely to report perpetrating sexual or economic violence (0.06 per cent in both cases,

corresponding to 120 and 97 men, respectively). More women than men reported having

perpetrated social, physical or psychological violence. These results must be interpreted

with great care because they are likely to be vastly under-reported. The results may also

exhibit reporting biases. Given the mismatch between the results in Table 45 and the

qualitative data, it is possible that women may have been more prone to report

perpetration of violence than men. With these caveats in mind, it is nonetheless noteworthy

that violence is perpetrated by both sexes. Table 46 provides further disaggregation of the

above results.

Table 46: Number and percentage of perpetrators reporting violence against same or

another sex

Social violence

Sexual violence

Physical violence

Psychological violence

Economic violence

Any domestic violence

Female perpetrators against male or female household members

Against men 176 (0.62) 3 (1.0) 90 (0.60) 106 (0.69) 30 (0.59) 248 (0.67)

Against women 193 (0.69) - 92 (0.61) 93 (0.61) 33 (0.65) 233 (0.63)

Total number 281 3 150 153 51 370

Male perpetrators against male or female household members Against men 71 (0.55) - 24 (0.69) 31 (0.49) 12 (0.24) 84 (0.50)

Against women 95 (0.73) 9 (0.90) 17 (0.49) 48 (0.76) 43 (0.86) 126 (0.75)

Total number 130 10 35 63 50 169


Table 46 shows that the largest share of domestic violence committed by women was

against men (0.67 per cent), but the difference between this and the share of violence

committed by women against other women (0.63 per cent) is small. Women were most

likely to commit violence against other women in the cases of social violence, physical

155

violence and economic violence (0.69, 0.61 and 0.65 per cent of women reported having

committed these forms of violence, respectively, against other women). Most of the

violence committed by men was against women (0.75 per cent, in comparison to 0.50 per

cent against other men). It is again important to keep in mind that these estimates may be

problematic, as discussed above. However, Tables 45 and 46 illustrate the level of

complexity and variety of domestic violence perpetration in Ghana.

Episodes of physical violence were also described as dynamic in that they could be driven

at first by one partner but ‘ended’ by another. This quote illustrates such dynamics, widely

reported at a regional level (but not nationally representative):

“The man gave the woman some blows on any part of her body that was available until she

fell down. The woman stood up and held the man by the neck, and the man began to give her

some blows until both of them fell down. They were so tired that they lay down till everybody

left the scene.” (FGD, Ashanti region, Adaase district: male and female group, aged

18+)

In some cases, the instance of physical violence followed a period of other forms of

violence. For example, some focus group discussions described how women struggled

with economic violence (such as men’s refusal to provide chop money) and that this violence

was compounded when women were ‘punished’ by their partner for failing to cook food

(itself a retaliation against the failure to provide money). Structural violence, and

pernicious poverty, might underlie this dynamic of intersecting forms of violence, as men

were not always able to secure wage labour and struggled with the pressure to ensure the

economic well-being of their household and its members (as is discussed later in the

report).

The qualitative data found, too, that some women withheld sex or shamed their partners

(through verbal abuse) when they did not earn enough money to support the running costs

of the household. This suggests an interlinked ‘web of violence’ that requires multiple

approaches for redress.

The study also collected information on how the perpetration of domestic violence varied

across socio-economic characteristics and geographical location of women and men. These

data are summarised in Tables 47 and 48, which suggest the following profiles:

The incidence of domestic violence perpetrated by women was highest among self-

employed and married women aged 25–29 years who lived in rural areas with no

or primary education and in the lowest asset quintile.

The incidence of domestic violence perpetrated by men was highest among men

aged 15–19 years who were employed and married, who lived in rural areas, were

in the lowest asset quintile but had primary or secondary education.

156

Table 47: Percentage of women reporting perpetration of domestic violence

Social Sexual Physical Psychological Economic Any

domestic

violence

Age group (years)

15–19

0.07

0.01

0.08

0.19

0.01

0.29

20–24 0.10 0.02 0.05 0.11 0.01 0.23 25–29 0.14 0.01 0.10 0.20 0.02 0.33 30–39 0.13 0.01 0.09 0.15 0.04 0.29 40–49 0.14 0.01 0.07 0.14 0.02 0.27 50–60 0.09 0.00 0.03 0.09 0.04 0.18

Employment

Self-employed

0.14

0.01

0.08

0.15

0.03

0.28 Employed 0.10 0.01 0.09 0.11 0.02 0.24 Not working 0.08 0.01 0.06 0.16 0.03 0.25

Marital status

Never married

0.04

0.00

0.09

0.17

0.02

0.25

Married or living together 0.12 0.01 0.08 0.14 0.03 0.27 Divorced/separated/widowed 0.13 0.01 0.05 0.15 0.03 0.26

Residence

Urban

0.11

0.01

0.07

0.13

0.02

0.25

Rural 0.13 0.00 0.08 0.17 0.03 0.29

Region

Western

0.14

0.01

0.09

0.16

0.04

0.29

Central 0.06 0.01 0.04 0.08 0.04 0.18 Greater Accra 0.10 0.02 0.07 0.11 0.03 0.21 Volta 0.09 0.01 0.11 0.31 0.02 0.39 Eastern 0.23 0.01 0.09 0.18 0.03 0.37 Ashanti 0.12 0.01 0.07 0.15 0.02 0.29 Brong Ahafo 0.08 0.00 0.06 0.11 0.02 0.20 Northern 0.14 0.00 0.13 0.18 0.03 0.33 Upper East 0.14 0.00 0.04 0.04 0.01 0.16 Upper West 0.15 0.04 0.04 0.08 0.02 0.20

Education level

None

0.13

0.01

0.08

0.15

0.03

0.28 Primary 0.14 0.00 0.08 0.16 0.04 0.28 Middle/JSS/JHS 0.12 0.01 0.07 0.15 0.02 0.27 Secondary 0.09 0.02 0.08 0.12 0.02 0.25 Technical 0.07 0.03 0.11 0.12 0.00 0.24 Higher 0.08 0.02 0.06 0.11 0.02 0.22

Asset quintile

Lowest

0.11

0.01

0.09

0.18

0.04

0.29

Second 0.14 0.01 0.07 0.17 0.03 0.28 Middle 0.10 0.01 0.07 0.16 0.02 0.27 Fourth 0.12 0.01 0.07 0.12 0.02 0.25 Highest 0.12 0.02 0.08 0.09 0.02 0.25

Total 0.12 0.01 0.07 0.15 0.03 0.27


157

Table 48: Percentage of men reporting perpetration of domestic violence

Social Sexual Physical Psychological Economic Any

domestic

violence

Age group (years)

15–19

0.07

0.06

0.09

0.22

0.04

0.35

20–24 0.12 0.12 0.07 0.18 0.04 0.32 25–29 0.12 0.10 0.07 0.14 0.04 0.33 30–39 0.10 0.04 0.05 0.14 0.07 0.27 40–49 0.16 0.04 0.04 0.12 0.09 0.33 50–60 0.11 0.02 0.04 0.14 0.04 0.25

Employment

Self-employed

0.12

0.05

0.05

0.13

0.05

0.28

Employed 0.12 0.07 0.07 0.15 0.08 0.34 Not working 0.11 0.08 0.06 0.19 0.03 0.32

Marital status

Never married

0.06

0.05

0.07

0.16

0.03

0.25

Married or living together 0.14 0.06 0.06 0.16 0.07 0.33 Divorced/separated/widowed 0.10 0.09 0.05 0.12 0.03 0.27

Residence

Urban

0.10

0.08

0.06

0.15

0.05

0.30

Rural 0.13 0.05 0.06 0.16 0.06 0.31

Region

Western

0.12

0.04

0.03

0.17

0.12

0.34

Central 0.09 0.09 0.08 0.13 0.03 0.26 Greater Accra 0.10 0.11 0.05 0.10 0.06 0.26 Volta 0.07 0.07 0.13 0.33 0.03 0.41 Eastern 0.24 0.05 0.08 0.17 0.05 0.40 Ashanti 0.13 0.07 0.05 0.16 0.04 0.33 Brong Ahafo 0.09 0.06 0.04 0.10 0.08 0.23 Northern 0.08 0.02 0.03 0.13 0.02 0.24 Upper East 0.13 0.00 0.07 0.05 0.05 0.23 Upper West 0.14 0.03 0.00 0.06 0.03 0.24

Education level

None

0.13

0.02

0.05

0.15

0.05

0.29

Primary 0.14 0.10 0.07 0.16 0.06 0.36 Middle/JSS/JHS 0.10 0.05 0.05 0.15 0.05 0.29 Secondary 0.10 0.10 0.07 0.16 0.05 0.32 Technical 0.09 0.08 0.07 0.15 0.05 0.25 Higher 0.14 0.05 0.05 0.15 0.08 0.31

Asset quintile

Lowest

0.10

0.04

0.06

0.16

0.06

0.31 Second 0.11 0.07 0.05 0.16 0.05 0.31 Middle 0.12 0.06 0.06 0.13 0.06 0.28 Fourth 0.13 0.07 0.07 0.16 0.06 0.31 Highest 0.13 0.08 0.07 0.14 0.05 0.30

Total 0.12 0.06 0.06 0.15 0.06 0.30


158

4.3.3. Domestic violence against children

In addition to the forms of domestic violence discussed in the sections above, the

qualitative data noted the occurrence of violence against children by adults who shared

the same home. The qualitative data, from all regions, found that women and men

employed physical violence to punish children, but there were different views on whether

physical violence, in particular, was an appropriate form of punishment for misbehaving

children. These views also differed depending on whether the person committing the

violence, or enacting the punishment, was male or female. For example, participants in a

focus group discussion with unmarried men and women (18–25 years old) in the Greater

Accra region said that only the father should beat children “because he is the breadwinner”.

In another focus group, also in the Greater Accra region, one respondent linked beating

with ‘correction’, saying, “fathers are responsible for correcting children through beating” (FGD,

Greater Accra region, Kukurantami district: male and female group, 18–25 years old).

Elsewhere, however, participants in the focus groups argued that this responsibility should

not be left only to men. In a focus group with married men (18–25 years old) in the Eastern

region, one participant asserted, “women should beat the children and the youth in the house

whenever they do something bad”. The GFLHS 2015 on domestic violence committed by

adults against their biological children found that the main perpetrator was the head of the

household (Table 49).

Table 49: Perpetration of domestic violence against individuals who reported domestic

violence and lived with their parents (in percentages)

Social

violence

Sexual

violence

Physical

violence

Psychological

violence

Economic

violence

Male head of household 88.0 0.0 100.0 87.0 82.0

Female head of household 87.0 100.0 63.0 70.0 87.0


Note: Values were reported by the household respondent and refer to children of the household head.

The qualitative data also highlighted that non-biological children may be at particular risk

of domestic violence inflicted by stepfathers, particularly when the mother is in a

vulnerable economic position. The following statement reflects this widespread finding

across regions:

“[Domestic violence] is a multi-faceted problem involving couples, children and at times in-

laws as well. Yes, at times it involves adults and their non-biological children. You

sometimes hear of men making advances towards the daughter of his wife who was brought

into the marriage. You can’t believe this, but it is true. When the man is supposed to treat

that child as his own daughter, he rather decides to make amorous advances towards her.”

(IDI, Upper East region, Garu Tempane district assembly, Garu opinion leader)

159

In addition to physical violence, controlling behaviour, such as fathers refusing to let their

daughters attend school, was noted as a form of domestic violence across all regions in the

qualitative data. The focus group participants were in significant agreement that this form

of violence was unacceptable. Given the nature of the qualitative data, it was not possible,

as with the quantitative survey, to gauge the full extent of these different forms of violence.

However, the data offer important insights into some of the underlying social norms,

beliefs and attitudes that individuals, families and communities hold around certain forms

of violence and their acceptability. These findings are discussed in further detail in the

section below.

4.4. Attitudes towards domestic violence in Ghana

The second research question asked how attitudes and social norms may influence the level

of acceptance of domestic violence in Ghana. Attitudes and beliefs – even if held at a

personal level – are usually shaped by interactions with people within the family; with

people in the immediate environment, including communities, neighbourhoods, social

networks and workplaces; and within economic or political structures. Understanding

how these reproduce or challenge gender inequality, gender norms and domestic violence

is crucial for designing interventions and policies that will effectively decrease the

incidence of domestic violence. For this reason, this component of the study sought to

explore, first, how individuals perceived and experienced domestic violence in their own

lives and families (the individual and community in the social ecological model) and,

second, how perceptions and experiences of individuals were informed by social norms

that may be collectively reinforced at the level of the community and broader social

contexts.

During the quantitative data collection, men and women were asked to discuss five

different scenarios that involved acts of violent behaviour. The aim of this exercise was to

understand to what extent these violent acts were perceived as acceptable, to what extent

victims of domestic violence were blamed for the violence (and under which

circumstances), and what respondents felt would be acceptable punishments for different

acts of violence. This information was complemented with qualitative data from focus

group discussions and interviews. To generate data to complement the survey, the focus

groups were also asked to respond to and discuss similar scenarios of social, physical,

sexual, psychological and economic violence. The focus group discussions and in-depth

interviews with opinion leaders, activists and legal experts explored in addition whether

and why certain types of behaviour were considered acceptable.

An important finding from the research centred on the variation in people’s understanding

of what constitutes: (i) violence in general; and (ii) domestic violence in particular. As a

result, the study offers a set of nuanced findings that complicate a straightforward reading

of domestic violence as per the Domestic Violence Act. Importantly, these nuanced

findings suggest that successful interventions need to be carefully tailored to raise

160

awareness around each of the categories of violence that are included under the general

term of ‘domestic violence’.

Given the scale of the qualitative and quantitative research, this section has been structured

to reveal findings for each of the five main types of violence explored in the study – namely:

social, physical, sexual, psychological and economic. While data have been disaggregated

against these forms of violence, it is important to recognise that individuals can experience

multiple forms of violence simultaneously; wherever possible, the multiplicity of domestic

violence is highlighted in the discussion below.

4.4.1. Attitudes towards domestic social violence

Both the quantitative and qualitative data collection made use of vignette scenarios to

understand individual and group-level attitudes towards domestic social violence. The

GFLHS 2015 included a range of vignette scenarios, and survey respondents were

randomly assigned scenario questions. In doing so, the survey generated findings about

different attitudes towards domestic violence by sex and other socio-economic

characteristics of the respondents. One of the first scenarios intended to find out about the

acceptability of jealous behaviour by a girlfriend or a boyfriend. Examples of jealous

behaviour in the vignette included: checking messages on their partner’s phone; following

them around; showing up at their workplace unannounced; or demanding to know their

whereabouts all the time. A second vignette scenario assessed respondents’ attitudes

towards women who stalked men, and compared those to attitudes towards men who

stalked women. Results for this analysis are included in Tables 50 and 51, respectively.

Overall, the majority of respondents thought that neither men nor women should ever

resort to the two forms of controlling behaviour hypothesised in the scenario questions.

The results for the first scenario were very consistent (Table 50), with around 70 per cent

of respondents on average across all socio-economic groups and sex answering that ‘he

should never do that’. On average, around 20 per cent of the sample across all groups

answered ‘I sympathise with him, but he should not do this’. There were interesting

patterns among those who answered ‘Yes, he has the right to do this’. Fifteen per cent of

the sample overall answered the question in this way, but men and individuals aged 15–

19 years, living in rural areas, in the Upper East and Upper West regions, with no education

and in the lowest asset quintile were more likely than others to answer ‘yes’ to this question

(Table 50). The reactions to the second scenario question were very similar (Table 51).

161

Table 50: Responses to first social violence scenario (in percentages)

Yes, he has

the right to

do this

I sympathise with

him, but he should

not do that

He should never

do that

Total

Sex

Male 12.1 19.5 68.4 100.0 Female 10.9 19.7 69.4 100.0

Age group (years)

15–19 14.4 19.8 65.8 100.0 20–24 8.6 17.8 73.6 100.0 25–29 9.6 18.3 72.0 100.0 30–39 13.0 18.4 68.6 100.0 40–49 11.0 22.7 66.3 100.0 50–60 11.8 21.4 66.8 100.0

Employment

Self-employed 11.5 19.3 69.1 100.0 Employed 11.7 20.6 67.7 100.0 Not working 10.7 19.7 69.6 100.0

Marital status

Never married 9.4 22.0 68.6 100.0 Married or living together 11.5 19.2 69.3 100.0 Divorced/separated/widowed 11.9 20.2 67.8 100.0

Residence

Urban 9.9 19.2 70.9 100.0 Rural 13.0 20.1 66.9 100.0

162

Yes, he has

the right to

do this

I sympathise with

him, but he should

not do that

He should never

do that

Total

Region

Western 7.0 25.6 67.5 100.0 Central 12.2 30.9 56.9 100.0 Greater Accra 8.6 13.3 78.1 100.0 Volta 4.5 32.3 63.1 100.0 Eastern 14.5 15.9 69.6 100.0 Ashanti 10.7 21.5 67.8 100.0 Brong Ahafo 11.7 17.7 70.6 100.0 Northern 7.5 10.7 81.9 100.0 Upper East 32.3 14.0 53.7 100.0 Upper West 49.0 8.5 42.5 100.0

Education level

None 14.3 21.3 64.5 100.0 Primary 12.2 21.0 66.8 100.0 Middle/JSS/JHS 9.5 19.5 71.0 100.0 Secondary 12.9 16.5 70.6 100.0 Technical 5.3 12.6 82.1 100.0 Higher 10.4 21.5 68.1 100.0

Asset quintile

Lowest 14.6 21.9 63.5 100.0 Second 12.0 18.4 69.6 100.0 Middle 8.0 22.6 69.4 100.0 Fourth 10.5 17.8 71.7 100.0 Highest 11.1 18.5 70.4 100.0

Total 11.4 19.6 69.0 100.0


163

Table 51: Responses to second social violence scenario (in percentages)

Yes, she has

the right to

do this

I sympathise with

her, but she should

not do that

She should never

do that

Total

Sex

Male 9.7 18.9 71.4 100.0 Female 10.1 19.3 70.6 100.0

Age group (years)

15–19 11.6 19.5 68.8 100.0 20–24 12.3 19.7 67.9 100.0 25–29 11.2 16.9 72.0 100.0 30–39 10.1 17.2 72.6 100.0 40–49 6.8 22.6 70.6 100.0 50–60 8.2 20.9 71.0 100.0

Employment

Self-employed 9.3 19.2 71.5 100.0 Employed 8.8 18.0 73.2 100.0 Not working 12.4 19.8 67.8 100.0

Marital status

Never married 15.1 15.9 69.1 100.0 Married or living together 9.1 20.3 70.6 100.0 Divorced/separated/widowed 10.2 16.9 73.0 100.0

Residence

Urban 9.7 18.9 71.4 100.0 Rural 10.2 19.4 70.4 100.0

164

Yes, she has

the right to

do this

I sympathise with

her, but she should

not do that

She should never do that

Total

Region

Western 9.4 21.9 68.7 100.0 Central 12.9 27.9 59.2 100.0 Greater Accra 9.5 18.5 72.1 100.0 Volta 1.1 29.3 69.6 100.0 Eastern 6.0 14.7 79.3 100.0 Ashanti 10.1 16.7 73.2 100.0 Brong Ahafo 11.1 14.9 74.0 100.0 Northern 12.6 14.9 72.5 100.0 Upper East 21.5 10.5 67.9 100.0 Upper West 28.4 20.9 50.7 100.0

Education level

None 12.6 17.2 70.2 100.0 Primary 9.4 18.4 72.3 100.0 Middle/JSS/JHS 8.6 19.2 72.2 100.0 Secondary 11.7 20.6 67.8 100.0 Technical 7.8 19.2 73.0 100.0 Higher 8.4 22.1 69.6 100.0

Asset quintile

Lowest 7.9 21.2 70.9 100.0 Second 11.4 17.8 70.8 100.0 Middle 12.4 20.1 67.5 100.0 Fourth 10.4 17.7 71.8 100.0 Highest 6.9 19.5 73.5 100.0

Total 9.9 19.1 70.9 100.0


165

As noted above, the qualitative research also included scenarios to understand the

participants’ attitudes towards social violence. These scenarios were discussed in all 20

focus groups. The findings from this component of the research were nationally

representative. They are discussed in detail below in relation to the quantitative findings,

and specific quotes are used selectively to illustrate certain findings.

Two complementary vignette scenarios on attitudes towards social violence were

discussed in the focus group discussions, whereby participants were asked to discuss: (i)

whether it was acceptable for men to prevent women from working (outside the home);

and (ii) whether it was acceptable for men to prevent their daughters from attending

school. The qualitative findings show that in most regions (with the exception of the three

northern regions) respondents found that it was unacceptable for men to prevent their

wives from working – particularly when the women contributed extra income to the

household. Results for all regions also showed widespread disapproval of men preventing

their daughters from attending school. Focus group discussions in the three northern

regions indicated higher variation in the response to whether or not it was acceptable for

men to prevent their partners from working. This finding is in line with a recently

published study on gender inequality in the Northern region (Alhassan and Odame, 2015),

where the authors outlined a set of household and contextual factors that worked against

government efforts to promote gender equality. These factors included limited availability

of household resources and poverty. Further socio-cultural practices, such as polygyny,

fostering, child marriage and boy-child preference were also shown to collectively

undermine women’s status relative to men, and to reinforce their economic dependence

on men in shared households. In the focus group discussions in the Northern region,

preventing women from working was viewed as acceptable because their work outside the

house would be at the cost of their domestic chores. The following quote illustrates this

view:

“[F]rankly speaking, juggling motherhood and wifely duties with a profession is no mean

joke. It’s really difficult. So if the man thinks his wife is suffering too much and tells her to

stay at home and take care of the house whiles he works to support them, I don’t see any

problems with that.” (FGD, Northern region, Sagnerigu district: female group, 18–50

years)

In contrast, the following dialogue, drawn from a focus group discussion with married

men (25–50 years old) in the Volta region, is an example of the prevailing response across

all regions to the scenario asking whether it is acceptable for men to prevent their

daughters from attending school:

Interviewer: Is it acceptable for fathers to stop daughters from attending school?

Respondent 7: It is not right for the father to deny his girl child education. Women are now

doing what some men cannot do. There are women who are taking care of their families.

Respondent 6: It is very wrong for the father to stop his girl child from attending school.

Women also need education to survive just like men.

Respondent 5: It is wrong to deny a girl child her right to education. Your daughter may

marry an educated person. The problem comes when the husband marries another woman

166

and the second wife happens to be educated, the man will definitely prefer the educated

woman to your daughter.

Respondent 4: It is not good for a man to ask his daughter to stop schooling. In this modern

era, women are equally doing what men can do.

Respondent 10: It is not right for a father to deny his girl child her right to education. The

reason is that capability is not based on gender.

Combined, the qualitative and quantitative findings drawn from the vignette scenarios

suggest that social violence was consistently viewed as unacceptable. However, social

violence between partners may also be shaped by gender norms – prevalent in the northern

regions – in which women were discouraged from engaging in paid labour. All

respondents consistently argued that it was unacceptable to prevent girls from acquiring

an education.

4.4.2. Attitudes towards domestic physical violence

As noted above, an important finding from the qualitative and quantitative data relates to

the variation in people’s understanding of what constitutes domestic violence. This section

offers a detailed analysis of attitudes towards domestic physical violence. The research

team again used scenario vignettes to elicit attitudes towards physical violence. The survey

introduced the following scenario: ‘After a marital dispute the wife refused to talk to her

husband, and refused to make his dinner and any other household chores. The husband

hit his wife, but without leaving a mark.’ The responses to this scenario are provided in

Table 52. Seventy-five per cent of male and female respondents found it completely

unacceptable for a man to beat his wife because she refused to talk to him and do

household chores after their dispute. In line with previous findings, this answer was most

common among respondents who were employed (82.3 per cent), living in urban areas

(80.2 per cent) and in the Greater Accra region (89.9 per cent), with higher education (83.8

per cent) and in the highest asset quintile (82.6 per cent).

167

Table 52: Responses to first physical violence scenario (in percentages)

Completely

unacceptable

Somewhat

unacceptable

Somewhat

acceptable

Completely

acceptable

Do not

know

Total

Sex

Male 75.0 13.4 4.8 6.4 0.4 100.0 Female 75.1 10.6 4.9 9.2 0.2 100.0

Age group (years)

15–19 74.3 7.7 5.6 11.0 1.3 100.0 20–24 70.5 14.7 4.3 9.7 0.7 100.0 25–29 77.4 10.1 4.9 7.6 0.0 100.0 30–39 75.5 11.9 4.9 7.6 0.0 100.0 40–49 75.5 10.4 5.5 8.3 0.3 100.0 50–60 75.3 14.2 4.0 6.2 0.3 100.0

Employment


Marital status


Residence

Urban 80.2 9.9 3.5 6.0 0.3 100.0 Rural 69.7 13.5 6.3 10.2 0.3 100.0

168

Completely

unacceptable

Somewhat

unacceptable

Somewhat

acceptable

Completely

acceptable Do not

know

Total

Region


Education level


Asset quintile


Total 75.1 11.7 4.9 8.1 0.3 100.0


169

Table 53: Responses to second physical violence scenario (in percentages)

Yes, fully Yes, partly Not at all Do not know Total

Sex

Male 48.2 32.7 18.5 0.6 100.0 Female 50.9 30.3 18.5 0.3 100.0

Age group (years)

15–19 55.9 20.5 22.3 1.3 100.0 20–24 49.6 30.5 19.2 0.7 100.0 25–29 49.9 29.3 20.6 0.3 100.0 30–39 50.3 30.9 18.6 0.2 100.0 40–49 47.4 33.4 18.8 0.4 100.0 50–60 48.4 38.8 12.6 0.2 100.0

Employment

Self-employed 52.4 30.6 16.8 0.3 100.0 Employed 40.9 36.4 21.8 0.9 100.0 Not working 49.7 29.3 20.5 0.4 100.0

Marital status

Never married 47.2 30.4 21.4 1.1 100.0 Married or living together 50.3 30.8 18.5 0.4 100.0 Divorced/separated/widowed 49.5 33.1 17.4 0.0 100.0

Residence

Urban 43.8 36.7 19.0 0.4 100.0 Rural 56.1 25.5 18.0 0.4 100.0

170

Yes, fully Yes, partly Not at all Do not know Total

Region

Western 57.6 22.6 19.8 0.0 100.0 Central 50.1 40.1 8.9 0.8 100.0 Greater Accra 25.9 52.6 21.2 0.3 100.0 Volta 39.6 36.4 23.2 0.8 100.0 Eastern 51.1 30.4 18.1 0.4 100.0 Ashanti 56.2 24.0 19.3 0.5 100.0 Brong Ahafo 56.5 21.5 21.5 0.5 100.0 Northern 64.4 22.0 13.7 0.0 100.0 Upper East 62.7 23.3 13.3 0.7 100.0 Upper West 67.1 14.1 18.7 0.0 100.0

Education level

None 59.0 27.3 13.7 0.0 100.0 Primary 54.4 28.2 16.5 0.9 100.0 Middle/JSS/JHS 49.8 31.4 18.1 0.6 100.0 Secondary 43.9 30.6 25.5 0.0 100.0 Technical 34.8 47.5 17.7 0.0 100.0 Higher 38.0 39.3 22.7 0.0 100.0

Asset quintile

Lowest 52.6 28.7 18.1 0.6 100.0 Second 58.2 25.6 16.1 0.2 100.0 Middle 51.4 26.9 21.4 0.4 100.0 Fourth 43.4 36.3 19.9 0.4 100.0 Highest 39.5 41.5 18.4 0.6 100.0

Total 49.8 31.2 18.5 0.4 100.0


171

The qualitative data from all regions similarly found that participants considered wife-

beating to be unacceptable. Wife-beating linked to women’s refusal to have sex was less

frequently mentioned in the focus group discussions, but when it was mentioned, both

female and male respondents iterated a consistent attitude, condemning men’s treatment

of women ‘as slaves’. In the excerpt below, for example, married men (aged 18–50) in the

Volta region explained why they believed that physical violence against women was

unacceptable. Their views were representative of the region’s findings (these issues were

not picked up in other focus group discussions), and they reflect the national-level data

from the interviews and other regional focus groups, condemning men as ‘weak’ for

beating women:

Respondent 2: No, it’s not acceptable because she’s his wife and deserves to be treated with

respect. She’s not a slave.

Respondent 7: No, it’s not acceptable. This is because the bible entrusts us to treat each

other with respect and love. After marriage, the two individuals become one, and so the man

ought to treat her as he would like to be treated.

Respondent 10: No, I also think it’s not acceptable because conflicts and disagreements are

integral parts of every relationship and so we can’t fall on beating to drive home our points.

Beating usually doesn’t solve anything so we must employ other means to get our wives to

agree with us.

Analysis of the qualitative data indicated that views about physical violence differed

depending on the age and relationship of the two (or more) people involved in the violent

incident or series of interactions. The data (in all regions) revealed in particular that

physical violence by parents against children was appropriate, and sometimes even

necessary. This statement, by a participant in a focus group discussion with married men

(aged 18–50) in the Brong Ahafo region, indicates this distinction about the acceptability of

violence against children by adults, on the one hand, and between men and women, on the

other hand:

Respondent 4: No, there is no justification whatsoever to beat your wife, because it is not a

father and child relationship, but you’re coming together as friends, and so whatever the case

may be, both of them should be able to iron their differences peacefully.

In research on attitudes to violence, predominantly in horizontal relationships (between

partners), scholars have found that the perceived culpability of the victim (i.e. whether

they are perceived to be at fault) plays a significant role in men and women’s perceptions

of the violence (Garcia-Moreno et al., 2005). Seeking to explore this factor in relation to

physical violence, the survey asked whether the respondents thought it was the woman’s

fault when she was beaten by her husband – in relation to the scenario discussed above.

The results of this analysis were striking and are presented in Table 53. Around half of the

sample on average across all groups answered that they fully agreed that the beating was

the woman’s fault – even though the majority of the respondents found the husband’s

behaviour unacceptable. This answer was more common among women: 50.9 per cent of

all women agreed that the woman in the scenario was at fault, compared to 48.2 per cent

of male respondents. The answer was also more common among self-employed

172

individuals (52.4 per cent), married (50.3 per cent), living in rural areas (56.1 per cent), with

no education (59.0 per cent) and in the lowest two asset quintiles (52.6 and 58.2 per cent,

respectively) (Table 53).

The survey also asked respondents under what circumstances wife-beating would be

acceptable. The five most accepted reasons for wife-beating are shown in Table 54, while

Table 55 expands this analysis across different socio-economic groups. The main reasons

why men and women justified wife-beating included disobedience and neglect of children:

11.2 per cent of male respondents and 17.2 per cent of female respondents mentioned that

wife-beating was acceptable if the wife had disobeyed the husband, and 10.2 per cent of

male respondents and 16.5 per cent of female respondents agreed that wife-beating was

acceptable if the wife neglected the children (Table 54). If the woman refused sex, a form

of violence largely condemned in all focus group discussions where this issue was raised,

5.7 per cent of men and 9.8 per cent of women thought this was an acceptable reason for

wife-beating (Table 55, Column 3).

Table 54: Percentage of respondents agreeing with wife-beating if the wife…

Male

respondents

Female

respondents

…disobeys her husband 11.2 17.2

…neglects the children 10.2 16.5

…goes out without telling her husband 7.3 13.2

…contracts a sexually transmitted disease 7.8 13.2

…argues with her husband 7.0 11.8


As noted above, women were consistently more likely to find wife-beating acceptable than

men (Tables 54 and 55). This finding echoes a similar result discussed in Takyi and Mann

(2009) and may imply that victims internalise social norms once they are in abusive

relations, as discussed in Section 2.

In line with findings discussed above, Table 53 shows that condoning wife-beating was

also more common, on average, among younger individuals, either self-employed or not

working, married, living in rural areas, with no education and in the bottom income

quintiles. There were also large regional differences: respondents from the three northern

regions were more likely to reveal attitudes supporting wife-beating than other regions.

173

Table 55: Percentage of respondents who think that a man has a good reason to hit his wife if:

1 2 3 4 5 6 7 8 9 10 11

Sex Male 4.6 11.2 5.7 2.0 5.6 0.4 7.8 7.3 10.2 7.0 2.4 Female 7.7 17.2 9.8 3.2 8.7 0.5 13.2 13.2 16.5 11.8 4.3

Age group

15–19 10.6

21.7 11.7 2.8 9.6 1.6 15.5 17.0 18.4 13.7 4.9 20–24 5.9 15.3 7.8 4.0 9.4 0.4 12.1 10.9 16.5 9.4 5.1 25–29 5.5 14.5 7.7 2.3 6.0 0.3 10.1 10.0 12.9 10.3 2.8 30–39 5.5 12.9 6.4 2.2 6.3 0.3 10.2 9.2 12.7 8.5 3.2 40–49 6.4 15.3 10.0 3.0 9.1 0.3 11.9 12.0 14.9 10.6 3.8 50–60 7.6 13.7 8.4 2.8 6.8 0.5 9.5 10.1 12.0 9.5 2.8

Employment

Self-employed 7.0 16.4 9.2 3.3 8.2 0.4 12.5 11.6 15.5 11.1 3.9 Employed 3.2 6.8 4.6 0.9 4.2 0.4 5.7 6.0 8.7 4.9 2.0 Not working 7.4 16.5 8.1 2.6 8.0 0.7 11.2 12.3 14.0 10.2 3.8

Marital status Never married 5.9 12.8 6.6 1.4 6.1 0.9 9.6 8.6 10.0 7.2 2.6 Married or living together 6.6 15.5 8.6 3.1 7.9 0.3 11.4 11.2 14.3 10.5 3.7 Divorced/separated/widowed 6.0 13.2 7.6 2.1 6.7 0.6 10.6 10.5 15.0 8.9 3.6

Residence Urban 3.4 9.3 4.8 1.3 4.7 0.4 7.2 6.9 9.9 6.0 1.6 Rural 9.6 20.5 11.7 4.2 10.5 0.5 15.1 15.0 18.3 13.9 5.6

174

1 2 3 4 5 6 7 8 9 10 11

Region

Western 4.4 14.1 8.2 0.9 10.0 0.2 8.9 13.0 13.2 8.3 4.0 Central 5.9 16.6 8.1 0.5 4.7 0.6 13.3 9.4 10.5 4.9 1.3 Greater Accra 1.2 3.1 1.7 1.0 2.6 0.1 3.0 2.6 3.9 1.7 0.8 Volta 4.9 7.3 5.2 3.3 4.7 1.6 9.9 7.2 9.5 8.0 4.5 Eastern 5.3 11.8 7.0 1.6 3.6 0.3 15.4 7.7 18.2 8.6 2.4 Ashanti 5.7 12.4 4.8 1.0 6.5 0.3 6.1 8.5 9.5 7.3 2.3 Brong Ahafo 3.6 16.1 7.3 2.7 7.7 0.5 10.1 8.6 12.4 12.8 2.6 Northern 14.8 42.7 24.4 11.3 15.0 0.6 21.3 28.5 38.6 29.1 6.7 Upper East 25.8 32.4 24.0 10.8 26.4 0.4 37.4 36.9 36.4 27.0 17.1 Upper West 31.5 38.6 29.4 8.9 27.8 0.4 34.9 30.2 37.4 29.3 16.7

Education level None 12.6 28.6 17.5 7.0 13.5 0.4 19.9 21.9 25.5 20.9 7.6 Primary 8.1 17.7 9.9 3.8 9.4 0.9 14.6 12.3 17.0 12.4 4.6 Middle/JSS/JHS 5.5 12.6 6.3 1.7 6.4 0.4 9.5 8.9 12.2 7.9 3.1 Secondary 3.1 8.2 3.5 0.9 4.9 0.3 6.2 6.7 8.5 4.8 1.2 Technical 2.0 4.3 5.1 0.6 4.3 0.1 5.2 2.6 3.9 1.3 0.7 Higher 1.6 4.6 2.5 0.3 1.8 0.1 3.4 2.9 5.1 2.5 0.5

Asset quintile

Lowest 10.6 22.5 13.4 4.7 11.2 0.9 18.5 17.7 20.5 15.5 6.0 Second 8.2 18.7 11.0 4.2 8.9 0.4 13.3 13.4 17.3 12.6 5.3 Middle 5.0 13.8 6.8 1.4 7.8 0.2 8.8 10.1 12.5 9.6 2.5 Fourth 4.4 10.1 4.6 1.7 5.6 0.3 7.8 6.7 10.6 5.8 2.1 Highest 2.4 5.9 3.1 0.7 2.5 0.4 4.6 4.1 6.4 4.0 0.7

Total 6.4 14.8 8.2 2.7 7.5 0.4 11.1 10.8 14.0 9.9 3.6


Note: 1 = She does not complete her household work to his satisfaction; 2 = She disobeys him; 3 = She refuses to have sex with him; 4 = She asks him whether

he has other girlfriends; 5 = He suspects that she is unfaithful; 6 = He finds out that she has been unfaithful; 7 = She contracts a sexually transmitted disease; 8

= She goes out without telling him; 9 = She neglects the children; 10 = She argues with him; 11 = She burns the food.

175

Fewer women in 2015 found wife-beating acceptable than in the 2008 DHS. Table 56

includes comparable questions asked in the GFLHS 2015 and in the Ghana DHS 2008. The

results show a reduction of 13.5 percentage points in the number of women justifying wife-

beating (from 36.6 per cent in 2008 to 23.1 per cent in 2015), and 8 percentage points in the

number of men condoning wife-beating for at least one of the reasons provided (from 21.8

per cent in 2008 to 13.8 per cent in 2015). Even though this attitude is still more prevalent

among women, it also seems to be changing more rapidly among women (the reduction is

larger among women).

Table 56: Percentage of respondents stating that wife-beating is acceptable if the

woman…

…refuses

to have sex

… goes out

without

asking

… neglects

children

… argues

with the

husband

… burns

food

At least

one

reason

2015 GFLHS

Men 5.7 7.3 10.2 7.0 2.4 13.8

Women 9.8 13.2 16.5 11.8 4.3 23.1

2008 DHS

Men 6.6 10.7 13.7 11.0 4.6 21.8

Women 12.3 22.3 26.3 21.3 8.3 36.6

Source: Ghana Family Life and Health Survey (GFLHS) 2015 and DHS 2008.

While the GFLHS 2015 focused predominantly on wife-beating, the qualitative focus group

discussions were able to explore in more detail men’s experiences of physical violence by

women – even though few numbers of respondents considered this a form of domestic

violence. This finding was regionally representative, with respondents in the Central,

Eastern, Volta and Greater Accra regions iterating the importance of ‘fairness’: just as

women should not be beaten, so too should women not beat men.

However, in the Northern and Upper East regions, the focus group discussions about

‘husband-beating’ – particularly among young men and women (aged 18–25) and among

married women (aged 18–50) – elicited a great deal of laughter, and sometimes even

disbelief. These responses to some of the group members’ accounts of men experiencing

physical violence were overlaid with gendered assumptions that these men were ‘weak’

and not able to maintain their superiority over women. For example, the excerpt below

from the Gbolo district in the Northern region offers a window into one such discussion,

where the possibility of a man being beaten by his wife was met with disbelief:

Respondent 2: If it is the woman beating the man, I would not separate them. I will pass by

because why should a man allow himself to be beaten by a woman?

Respondent 7: I will go and separate them, but if I could, I will support the man to sack the

woman/wife because it is disrespectful. Why should you beat up your husband?

Respondent 6: I will separate them.

176

Respondent 4: I will not mind them. I will pass by because why a woman should beat her

husband is a taboo in our community.

(FGD, Northern region, Gbolo district: male and female group, 18–25 years)

However, sometimes the laughter had a serious undertone, revealing discontent with

gender norms among the respondents, particularly in the Northern and Upper East

regions. In some groups, the question revealed a desire for revenge by women who had

themselves experienced multiple forms of domestic violence. In this case, this finding

suggests that violence may be deemed ‘acceptable’ when it constitutes, or is viewed to

constitute, a form of retribution. This sentiment, while widely reported mostly in the

Northern and Upper East regions, is powerfully captured in this discussion among

married women (aged 18–50).

Respondent 1: I’ll stand by and watch as the woman beats him, I’ll neither call for help nor

try to separate them because he’s beaten his wife and so he should also have a taste of what

it takes to be beaten.

Respondent 2: If I meet such a scenario, I’ll simply let the woman beat the man because he’s

a wife-beater and so he must be taught a lesson in respecting women, so I’ll cheer the woman

on.

Respondent 7: As for me, I’ll help the lady to give the man a good beating. I’ll even referee

the fight.

(FGD, Northern region, Sagnerigu district: female group, 18–50 years)

This dynamic, however, did not apply to men’s views of the acceptability of violence

against women.

The qualitative and quantitative data have also allowed the research team to examine

attitudes towards physical violence against children. This form of violence was

predominantly considered unacceptable. However, some study participants expressed

certain contexts when physical violence against children was acceptable or even desirable.

The qualitative data revealed how, across all regions, parents often struggled to discipline

their children when they misbehaved. Focus group discussions in all ten regions revealed

that both older and younger respondents of both sexes agreed that there were certain

circumstances when physical violence against children was acceptable. This finding

suggests that not only do gender norms shape the perceived acceptability of certain forms

of violence between partners, but that age is an important feature for distinguishing what

kinds of physical violence are and are not appropriate. This finding is poignantly

illustrated in the following excerpt from a focus group discussion in the Northern region,

which represents views expressed across the whole country:

Respondent 2: Yes, it is acceptable somehow because if they do what is not right, the father

has to talk to them and if they persist, then the father can beat them.

Respondent 4: Yes, it is acceptable somehow because the man has to talk to the children for

the third time and if they persist then the man can beat them.

Respondent 7: Yes, but it is very necessary for the men to talk to the children for the first

time and if they prove stubborn, then they have to beat them.

177

(FGD, Northern region, Gbolo district: male group, aged 50+)

To explore the circumstances under which physical violence against children was deemed

acceptable, the GFLHS 2015 included a scenario question asking respondents whether they

would intervene if they witnessed an act of physical violence against a child. Ninety per

cent of respondents said they would intervene through a variety of means, including their

own direct intervention or by approaching religious authorities, community leaders, a

relative of the child, a neighbour or other community members. 53 per cent of men and

50.8 per cent of women in the GFLHS 2015 mentioned that they would intervene directly

by going to check what happened, trying to rescue the child themselves or trying to stop

the dispute.

Physical punishments of children seemed to be at least partially acceptable if they intended

to discipline the child when non-physical attempts were ineffective. The GFLHS 2015

asked respondents whether they fully agreed, somewhat agreed, somewhat disagreed or

fully disagreed with the following statement: ‘To bring up, raise or educate a child

properly, the child sometimes needs to be physically punished.’ 50 per cent of women and

43.3 per cent of men fully agreed with this statement (Table 57). Younger respondents were

more likely to fully disagree with the physical punishment of children (over 24 per cent

among those aged 15–24 years, compared to 19.9 per cent among respondents aged 50–60

years), as were employed respondents, living in urban areas and in the Western and

Ashanti regions, and those with higher levels of education and assets (Table 57). There has

been, however, a significant change in attitudes towards physical violence against

children: in 1998, 8 per cent of respondents thought it was never acceptable to beat a child

(Coker-Appiah and Cusack, 1999). In 2015, this number has risen to 25.3 per cent among

men and 18.8 per cent among women (Table 57).

178

Table 57: Views about statement ‘To bring up, raise or educate a child properly, the

child sometimes needs to be physically punished’ (in percentages)

Fully

agree

Somewhat

agree

Somewhat

disagree

Fully

disagree

Do not

know

Total

Sex Male 43.3 18.9 12.3 25.3 0.1 100.0 Female 50.6 18.7 11.9 18.8 0.0 100.0

Age group (years) 15–19 48.1 15.2 11.8 24.8 0.0 100.0 20–24 45.7 17.6 12.3 24.3 0.0 100.0 25–29 47.6 19.7 10.8 22.0 0.0 100.0 30–39 48.1 18.3 12.4 21.1 0.1 100.0 40–49 49.6 18.9 12.7 18.7 0.1 100.0 50–60 46.4 21.4 12.2 19.9 0.0 100.0

Employment

Self-employed 49.2 19.1 12.7 19.1 0.0 100.0 Employed 44.2 17.8 12.2 25.5 0.3 100.0 Not working 46.4 18.8 10.4 24.4 0.0 100.0 Marital status

Never married 50.8 15.6 11.6 22.1 0.0 100.0 Married or living together 47.0 19.2 12.1 21.7 0.0 100.0 Divorced/separated/widowed 48.5 19.0 12.3 20.2 0.0 100.0 Residence

Urban 45.8 19.5 11.4 23.2 0.0 100.0 Rural 49.7 18.0 12.8 19.5 0.1 100.0 Region

Western 50.6 10.9 8.5 30.0 0.0 100.0 Central 46.5 17.7 22.1 13.7 0.0 100.0 Greater Accra 49.2 20.8 10.3 19.8 0.0 100.0 Volta 40.7 32.6 13.4 13.3 0.0 100.0 Eastern 67.2 18.2 6.7 7.9 0.0 100.0 Ashanti 45.4 8.8 8.2 37.3 0.2 100.0 Brong Ahafo 42.1 23.0 13.3 21.6 0.0 100.0 Northern 31.2 28.0 24.7 16.2 0.0 100.0 Upper East 54.2 18.7 10.1 17.1 0.0 100.0 Upper West 63.8 21.8 6.8 7.6 0.0 100.0 Education level

None 45.2 23.3 15.3 16.2 0.0 100.0 Primary 53.7 18.0 11.5 16.8 0.0 100.0 Middle/JSS/JHS 49.2 17.0 12.0 21.6 0.1 100.0 Secondary 44.8 17.3 9.5 28.5 0.0 100.0 Technical 38.6 25.5 12.1 23.8 0.0 100.0 Higher 45.0 18.4 11.0 25.7 0.0 100.0 Asset quintile


Total 47.7 18.8 12.1 21.4 0.0 100.0


179

4.4.3. Attitudes towards domestic sexual violence

The GFLHS 2015 revealed nuanced attitudes of different individuals and social groups

towards domestic sexual violence, with many exhibiting a higher degree of acceptance of

some dimensions of sexual violence than other forms of violence. The research team

explored attitudes towards domestic sexual violence by asking respondents their views in

relation to two statements: (1) ‘If a woman wears revealing clothes, it is okay for men to

make some sexual comments as she walks in her neighbourhood’; and (2) ‘If a woman

wears revealing clothes, it is her fault if she is raped.’ Tables 58 and 59 show, respectively,

the reactions of the respondents to these two statements.

Findings show that a very substantial part of the sample – 36.7 per cent of men and 40.7

per cent of women – fully agreed with the first statement (Table 58), without much

variation in responses across socio-economic groups. Individuals with higher levels of

education, in the top asset quintile and living in the Eastern and Ashanti regions and in

urban areas were, however, more likely to fully disagree with the first statement (Table 58,

Column 4), even though ‘fully agree’ was the most common answer across most socio-

economic groups.

More than three in every five respondents (61.6 per cent) fully agreed that the woman was

to be blamed for rape if she wore revealing clothes. As with other forms of violence, a

higher proportion of women than men displayed attitudes condoning this: 65.3 per cent of

women interviewed agreed with the second statement, compared to 56.2 per cent of men.

Respondents from rural areas were more likely than those in urban areas to fully agree

with the second statement (65.8 and 57.7 per cent, respectively). The data show no other

substantial differences across other group divisions. Employed individuals with higher

levels of education, in the top asset quintile and living in urban areas were more likely to

fully disagree with the second statement (but ‘fully agree’ was the most common answer

among most socio-economic categories).

180

Table 58: Views about statement ‘If a woman wears revealing clothes, it is okay for

men to make some sexual comments as she walks in her neighbourhood’ (in

percentages)

Fully

agree

Somewhat

agree

Somewhat

disagree

Fully

disagree

Do not

know

Total

Sex

Male 36.7 17.2 16.6 28.9 0.6 100.0

Female 40.7 16.9 14.4 27.4 0.6 100.0 Age group (years)

15–19 40.4 15.0 13.6 28.6 2.4 100.0 20–24 40.6 14.8 13.7 30.8 0.1 100.0 25–29 38.9 17.4 14.9 28.5 0.3 100.0 30–39 38.2 17.5 16.3 27.5 0.5 100.0 40–49 40.8 17.4 15.1 26.0 0.6 100.0 50–60 36.9 18.3 16.4 27.7 0.7 100.0

Employment


Marital status

Never married 36.1 17.5 14.4 29.9 2.1 100.0 Married or living together 38.9 17.0 16.1 27.6 0.4 100.0 Divorced/separated/widowed 41.2 17.1 13.0 28.4 0.3 100.0 Residence

Urban 38.2 17.0 15.0 29.1 0.7 100.0 Rural 40.0 17.1 15.6 26.8 0.5 100.0

Region



Asset quintile


Total 39.1 17.0 15.3 28.0 0.6 100.0


181

Table 59: Views about statement ‘If a woman wears revealing clothes, it is her fault if

she is raped’ (in percentages)

Fully

agree

Somewhat

agree

Somewhat

disagree

Fully

disagree

Do not

know

Total

Sex

Male 56.2 11.0 8.8 23.7 0.4 100.0 Female 65.3 9.5 6.7 18.2 0.3 100.0 Age group (years)

15–19 63.5 9.8 8.5 17.6 0.7 100.0 20–24 64.0 9.2 5.8 20.5 0.5 100.0 25–29 62.9 11.0 6.9 18.5 0.7 100.0 30–39 61.7 11.3 7.8 18.9 0.2 100.0 40–49 59.8 9.2 9.0 21.9 0.1 100.0 50–60 58.8 8.6 7.5 24.9 0.1 100.0 Employment

Self-employed 63.3 9.6 8.3 18.6 0.2 100.0 Employed 54.8 11.7 6.7 26.3 0.6 100.0 Not working 62.3 10.1 6.4 20.7 0.5 100.0 Marital status


Residence

Urban 57.7 10.1 8.8 23.0 0.5 100.0 Rural 65.8 10.0 6.3 17.7 0.2 100.0 Region


None 59.5 12.4 8.8 18.9 0.4 100.0 Primary 68.9 9.4 6.1 15.5 0.2 100.0 Middle/JSS/JHS 65.4 8.3 8.1 17.7 0.4 100.0 Secondary 59.6 10.3 6.1 23.7 0.3 100.0 Technical 58.5 7.3 7.7 26.5 0.0 100.0 Higher 43.8 14.0 7.5 34.4 0.3 100.0 Asset quintile


Total 61.6 10.1 7.6 20.4 0.3 100.0


182

There was limited discussion about the incidence of sexual violence in the focus group

discussions and individual interviews. Qualitative data on attitudes towards sexual violence

revealed (in the groups that discussed this form of violence), however, that this form of

violence was unambiguously perceived as unacceptable, across all regions. This view is

illustrated, for example, in the following focus group discussion with married men (18–50

years) in the Brong Ahafo region, in the Ahafo Ano North municipal district:

Respondent 3: No, it’s not acceptable. Sex is based on understanding and affection and is only

enjoyable if both partners willingly participate. The woman may not be in the mood so if she’s

not in the mood, how do you enjoy it? Besides how do you beat your partner and have sex with

her the next moment? It’s really cruel.

Respondent 4: No it’s not acceptable. Marriage is based on love and not sex; sex is just one

aspect of marriage so it shouldn’t dictate how you treat your partner. The man should be able to

understand and empathise with the woman in those times that she’s unable for some reason to

give him sexual gratification. There would definitely be other opportunities.

The previous section discussed how patterns of domestic sexual violence differed from other

types of domestic violence. In particular, the results revealed that experiences of sexual

violence were more common among employed, wealthier and more educated individuals

living in urban areas, particularly among men. The research team postulated that this finding

may have emerged because these groups are also more likely to have reported experiences of

sexual violence. The findings in this section go some way towards supporting this. First, there

was very limited discussion about sexual violence in the qualitative data. Second, most survey

respondents showed high levels of acceptance of behaviours defined by sexual violence.

Taken together, these findings suggest that sexual violence may be understood largely as

something that takes place outside domestic relations. Perceptions of what may constitute

sexual violence within domestic relations seem very limited – or hidden. This latter hypothesis

is partially substantiated by the additional finding that larger percentages of female

respondents accept forms of sexual violence, suggesting that women may internalise gender

norms and norms about the acceptance of certain types of behaviour, as discussed in Section

2 and found in other studies.

4.4.4. Attitudes towards domestic psychological violence

As discussed in the sections above on attitudes towards domestic social and physical violence,

the qualitative and quantitative data found that there were circumstances where certain forms

of domestic psychological violence were viewed as defence mechanisms and as a way of

women claiming some limited power in highly unequal domestic partnerships. This finding

was nationally representative in the qualitative data. The following quote, from a participant

in a focus group discussion among young men and women aged 18–25 years in the Eastern

region, captures this finding succinctly:

“Most of the fights are physical but some are verbal. Most of the verbal fights are from the women

and they use them as a defensive mechanism.”

183

Echoing the findings from the qualitative data, data from the GFLHS 2015 revealed that 31 per

cent of acts of insulting, belittling or humiliation of male heads of households (in private or in

front of other people) were perpetrated by their female spouse, partner or former partner. The

remainder were perpetrated by other relations. The incidence of psychological abuse of the

same type against female household heads was lower, around 25 per cent, according to the

GFLHS 2015. The qualitative data provided important explanations for these levels of

incidence. Notably, focus group participants often perceived the exercise of psychological

pressure in the form of verbal abuse as a means of defence and a coping strategy (for women,

in particular) to ward off physical violence, or to claim some autonomy or agency in a context

where women may not have the ability to exert much power beyond the use of words (insults),

their bodies (withholding sex) or their housework (not cooking).

The qualitative data also offered some striking (but not nationally representative) findings that

reveal how men also use psychological violence as a form of retribution. For example, in a

mixed focus group discussion with youth in Kukurantumi (FDG, Eastern region, East Akim

district: male and female group, 18–25 years), a respondent illustrated this finding as follows:

“Let us assume that my wife has offended me but I don’t want to beat her. I can abuse her by not

talking to her for a couple of months.”

In a focus group discussion among young men and women (aged 18–25) in the Ashanti region,

in Obuasi district, a respondent outlined an additional factor that may lead to abuse of women

by men as a form of retaliation:

“Women who are rich or self-made. Who doesn’t respect or appreciate their husbands, especially

if the man is not rich like her, a case in point is when the woman doesn’t appreciate the man’s

gifts, the man feels unappreciated resulting in drunkenness, psychological abuse, etc.”

To explore attitudes towards psychological violence further, the GFLHS 2015 included the

following scenario: ‘A woman repeatedly makes demeaning comments to her husband about

him not making enough money, and not being able to provide for a nice home and gifts for

her. She constantly compares him to his brother, who is a successful businessman.’

Respondents were asked whether this was a scenario that they were familiar with. Nearly half

(48.8 per cent) of respondents stated that this situation was ‘very common’, and 22.9 per cent

that it was ‘somewhat common’. The survey then repeated this scenario with the variation that

the woman was beaten by her husband following the remarks and reported the incident to the

police. Just over half of the respondents (55.8 per cent) agreed that she should have reported

the abuse to the police. Of a total of 1,098 respondents (412 men and 686 women) who thought

the woman should not have reported the incident to the police, 25.2 per cent of men and 32.1

per cent of women mentioned that she should not have reported the beating to the police

because the woman was “the one to blame” for the incident. This finding is in line with the

discussion in the previous sections, where attitudes condoning domestic violence seem more

prevalent among women.

184

4.4.5. Attitudes towards domestic economic violence

Throughout the qualitative data, respondents across all regions expressed strong attitudes

against domestic economic violence. However, the qualitative data drew out interesting

paradoxes. While most forms of domestic violence against women were generally seen as

unacceptable, the data have also shown that men – but not women – were seen to be the main

providers of material well-being to the household and its members. This argument was

forcefully made, in particular, by married women (aged 18–50) in the focus group discussions

that were conducted across all ten regions of the country. The flipside of the argument, which

was made less often by married men, was that the pressure to secure a job and be the sole

breadwinner was also a form of economic violence exerted by women who had not been

restricted from earning an income but who had chosen not to. The following excerpt illustrates

some of the different attitudes women expressed about economic violence. In this particular

conversation, the women articulated one major form of economic violence that was expressed

across the qualitative data – namely, men’s refusal to give women chop money for household

costs:

Facilitator: Is it acceptable if a man refuses to give chop money to his wife?

Respondent 8: No. From the Bible, God said husband should take care of their wives for that

reason any man who marries and doesn’t take care of the woman is not doing right.

Respondent 5: It is not the right thing to do. You have to give her chop money.

Facilitator: Can you explain your answer?

Respondent 2: Because you the man brought the woman to your house so every morning, you

must give her chop money. If you refuse to give me chop money, there will be a fight in the house.

Respondent 3: No. Because he brought me to live with him, so he must take care of me well.

The following extract from a focus group discussion with married women (18–50 years) in the

Brong Ahafo region, in the Ahafo Ano North municipal district, similarly illustrates how

gender norms – and expectations that men should be the ‘breadwinner’ – affect women’s

perception of economic violence. The views expressed by the respondents below were

common across most of the sample and across all regions:

Respondent 3: No, it is not acceptable because it is the man’s responsibility to take care of the

whole family. Since the man cannot starve himself throughout the whole day, he should not also

attempt to starve the rest of the family members. The man cannot say that I cannot refuse giving

chop money to the family. In some instance where the man is not financially sound, he can tell

the woman politely to support him if she has but not to intentionally refuse to give the wife chop

money. It is only a wicked person who will refuse to give the woman chop money.

Respondent 4: It is a big no, because the man is the breadwinner of the family and hence his

responsibility to take care of the family. If for one reason or the other the man does not have

money, he can call the woman to support for that period but not to intentionally refuse to give

her chop money. Considering the way our economy is now, the women are also liable to support

but the responsibility lies on the man to take care of the household.

Respondent 5: It is not acceptable. There are times that the man may genuinely not have money

due to loss of job. Within that period the man can ask the woman to support the house. But, if

the man has money, it not acceptable to refuse to give the woman chop money.

185

Respondent 8: No, it is not acceptable because considering the main responsibilities of a man

at home, providing food is one of their core responsibilities. So if by God’s grace you have money

I don’t see the reason why you should refuse to give chop money. What will happen then is that,

for days that you will refuse to give chop money we will not also leave food for you to come and

eat.

Men felt the strain of these expectations. This excerpt from an interview with an activist in the

Ashanti region describes this view, which was expressed across discussions with men:

“Men tend to experience violence if they are not playing their role well. I am talking about roles

because in Africa, we expect the man to provide. At least make the petty provision, not the

essential. At least to maintain the family. So if you are not maintaining the family, definitely,

the lady turns hostile. When the lady becomes hostile, it rolls unto you. The lady will not allow

you to make love with her, and then she will try to fight you, especially if you are not strong.

And then that is where the men turn to suffer. At times some of the men come here to complain

about the women slapping them, but because of the scriptures and the fact that they are God

fearing, they cannot do anything.”

The denial of chop money also came out very strongly in the quantitative survey. It was the

single most important act of economic violence mentioned against women and men (Tables

23–26): 11.6 per cent of women and 5.7 per cent of men reported having been denied chop

money in the 12 months prior to the survey, despite believing there was enough money

available for those expenses.

4.5. Determinants of domestic violence in Ghana

This section discusses a number of factors the study found to be associated with different

forms of violence against women, men, girls and boys, including where people live, and the

social and economic characteristics of individuals, households and communities. In line with

the sections above, the analysis combined insights from both quantitative and qualitative data.

The sections also include information from multivariate regression analysis, as described in

the methodology section, which has attempted to identify the main determinants of domestic

violence in Ghana in 2015.

To conduct the regression analysis, the research team created a variable for each category of

domestic violence (social, physical, sexual, psychological and economic). This variable takes

the value 1 if the person reported having been exposed to any act of violence within each of

these broad categories in the 12 months prior to the interview; otherwise the variable takes the

value 0. In line with the study’s conceptual framework, the incidence of domestic violence was

regressed on a set of variables reflecting individual, household and community factors, as well

as a set of regional variables.25

25 Each of the ten regions of Ghana is represented by a dummy variable with value 1 if the respondent lives

in the region, and 0 otherwise. The inclusion of regional variables – or regional fixed effects – allowed us to

account for some of the unobserved variation of domestic violence across regions that would otherwise be

186

The variables used in the regressions were derived from the conceptual framework discussed

in Section 2, as follows:

Intergenerational dimension of domestic violence: measured using a binary variable

indicating whether the respondent witnessed each type of domestic violence in their

home during childhood.

Marital status: measured with a variable indicating whether the respondent: (i) was not

married ever; (ii) was in a monogamous relationship; (iii) was in a polygamous

relationship; or (iv) was divorced, separated or widowed.

Socio-economic status: captured by an asset ownership index. The assets included were:

radio, TV, mobile phone, bed, other furniture, cooking utensils, washing machine, fan,

air conditioner, refrigerator, iron, sewing machine, kerosene stove, gas/LPG/biogas

stove, electric stove, open stove, dish/cable TV, wall clock, landline telephone, digital

camera, tablet, computer and mosquito net. The research team chose to measure socio-

economic status through ownership of assets, as this measure is generally recognised

as including less measurement error than income (Deaton, 1997).

Education: measured using a variable that indicated the different levels of schooling

attained by each respondent.

Empowerment: following the existing literature (see review in Justino et al., 2012),

empowerment was measured using data on decision-making. The research team

constructed a decision-making index to identify the involvement of the respondent in

making decisions about the following domains of household life: food purchases, daily

purchases, major purchases and health care (for self).

Patriarchal norms: following the discussion on attitudes towards domestic violence in

the previous sections, the prevalence of patriarchal norms was measured using a

patriarchal gender norm index,26 a standardised index of acceptance of wife-beating27

and an index of tolerance of women’s sexual autonomy.28

included in the error term. Since observations are clustered at enumeration area level, the regressions show

corrected standard errors. 26 This index was constructed based on the number of the following statements for which the respondent

fully or somewhat agrees: ‘A good wife obeys her husband even if she disagrees with him’, ‘Family problems

should only be discussed with people in the family’, ‘To bring up, raise or educate a child properly, the child

sometimes needs to be physically punished’, ‘If a women wears revealing clothes, it is her fault if she is raped’

and ‘It is allowed for men to stare at women they don’t know’. The index was standardised to have a mean

of 0 and a standard deviation of 1. 27 This index correspond to the number of the following cases for which the respondent agrees that a husband

has a good reason to hit his wife: “she does not complete her household work to his satisfaction”“, “she

disobeys him”, “she refuses to have sex with him”, “she asks him whether he has other girlfriends”, “he

suspects she has been unfaithful”, “he finds out she has been unfaithful”, “she contracts a sexually

transmitted disease”, “she goes out without telling him”, “she neglects the children”, “she argues with him”

and “she burns the food”. The index was standardised to have a mean of 0 and a standard deviation of 1. 28 This index corresponds to the number of the following cases for which the respondent agrees that a married

woman can refuse to have sex with her husband: “she does not want to”, “he is drunk”, “she is sick”, “he

mistreats her”, “she is menstruating”, “he does not want to use a condom/contraception” and “she finds out

that he has other girlfriends”. The index is standardised to have a mean of 0 and a standard deviation of 1.

187

Exposure to other forms of violence: measured using an index that included how many of

the following types of violence the respondent had witnessed in their community: theft

or robbery, vandalism, arson, kidnapping or abduction, extortion, bullying, political

violence, highway attacks, harassment, rape, defilement and seduction.29

In addition to the variables above, the pre-analysis of the qualitative and quantitative data

suggested the inclusion in the regression analysis of an additional variable – drinking habits of

the respondents, measured by a binary variable taking the value 1 if the respondent reported

drinking, even occasionally. The regressions control also for the region of residence and

whether the household lives in urban or rural areas.

The regression results are presented in tables F1–F4 in Appendix F, and discussed below, in

conjunction with the findings of the qualitative data.30

4.5.1. Age and intergenerational effects

The literature discussed in Section 2 revealed that experience of domestic violence is often

found to be closely correlated to the age of the victim. In particular, younger women are

generally more likely to report domestic violence than older women (Aizer, 2011; Garcia-

Moreno et al., 2005). The data collected in this study for Ghana are in line with this

international pattern: younger women – and men – were found to be more likely to have

experienced any form of domestic violence than older age groups, as discussed in

subsection 4.1.

The results of the multivariate regressions confirmed these results. Older women were less

likely to report experiencing domestic violence than younger women (Table E1, in

Appendix E). Women aged 50–60 years were 84 per cent less likely to have experienced

domestic violence than women aged 15–19 years (Table E1, in Appendix E). The same

figures for physical, sexual, social and economic violence, respectively, were 99 per cent,

98 per cent, 89 per cent and 73 per cent. Even women aged 20–24 years were significantly

less likely than women aged 15–19 years to experience all forms of domestic violence,

except economic violence, for which there is no statistically significant difference (Table

29 The sample mean for this variable is 0.56, and the median is 0. This indicates that violence tends to

concentrate in a few areas. We, therefore, use a binary variable taking the value 1 if the index of violence

exposure is positive, and 0 otherwise. Similar results in the regression analysis were obtained when we used

the continuous index. 30 The tables in Appendix E display the odds ratios associated with every risk factor, alongside its standard

errors. The odds ratio shows how much more (or less) the odds are of experiencing domestic violence when

a certain risk factor is present, in relation to respondents who did not experience the risk factor. An odds

ratio greater than 1 means that the variable increases the risk of domestic violence. As an example, an odds

ratio of 1.5 for individuals with no education indicates that the odds of experiencing domestic violence are

50 per cent higher among individuals with no education in relation to those with education. An odds ratio

of 2 for individuals with no education indicates that the odds of experiencing domestic violence are 2 times

(100 per cent) higher among individuals with no education than those with education. If instead the odds

ratio is 0.2, this indicates that people with no education are 80 per cent less likely than people with other

levels of education to experience domestic violence.

188

E1, in Appendix E). The odds of men experiencing psychological, physical, social or

economic domestic violence were also reduced as age increased (Table E2, in Appendix E),

though the statistical significance of this association is more variable than the results shown

for women. The odds of men being victims of sexual violence did not vary significantly

with age.

The literature reviewed in Section 2 also discussed how domestic violence is typically

related to strong intergenerational effects (Clark et al., 2010; Gallegos and Gutierrez, 2011;

Bobonis et al., 2013; Pollak, 2004), a result that was also revealed in the Ghana 2008 DHS

study (GSS et al., 2009). This study found similar results for Ghana in 2015. Except for

sexual violence (which may be more hidden from the view of children than other forms of

domestic violence), women who witnessed any form of domestic violence at home as

children were significantly more likely than women who did not witness violence in their

childhood to have been exposed to the same type of domestic violence in the 12 months

prior to the survey (Table E1, in Appendix E). The variations were large: women who

witnessed psychological violence as a child were 2.2 times more likely to be exposed to

psychological violence than women who did not witness this form of violence during their

childhood. The corresponding figures for social and economic violence were almost two

and three times higher. The effect for physical violence was more modest but remained

large: women who witnessed physical violence as a child in their homes were 73 per cent

more likely to be exposed to physical violence than women who did not witness this form

of violence during their childhood.

Men who witnessed social or economic violence in their homes as children were 2.2 times

more likely to have experienced these forms of violence than men who did not witness

these forms of violence in their childhood. Men who witnessed psychological violence as

a child were 80 per cent more likely than men who did not witness psychological violence

in their homes during childhood to have experienced this form of violence as an adult in

the 12 months prior to the survey. The effect was not statistically significant for physical or

sexual violence (Table E2, in Appendix E).

Witnessing violence as a child, therefore, has important generational effects, as these

children were more likely than others to have experienced domestic violence themselves

as adults. The effect also applies to the perpetration of domestic violence. According to the

GFLHS 2015, about 28 per cent of women and men who admitted committing social

violence witnessed social violence as children, compared to 14–15 per cent of those women

and men who did not experience violence as children. Similarly, among men who

witnessed sexual violence as children, almost 30 per cent committed sexual violence over

the 12 months prior to the survey, compared to 2.6 per cent of men who did not experience

this form of domestic violence during their childhood. Across psychological, physical and

economic violence, rates of perpetration are between two and three times higher for

women and men who witnessed domestic violence as children than among individuals

who did not experience domestic violence during their childhood.

189

4.5.2. Socio-economic effects

Section 2 discussed in detail how domestic violence was generally found to be associated

with low levels of income, wealth and education in many parts of the world. These socio-

economic factors were found to be some of the most important determinants of domestic

violence in Ghana in 2015. The vast majority of the research participants identified fights

over money, assets or property as the leading determinants of domestic violence across all

regions. Economic tensions were mentioned as leading to multiple forms of domestic

violence among adults – including between partners, siblings and in-laws. Other factors

reported frequently included education, poverty and the timing of certain economic

events. These findings are discussed in turn below.

Assets, property and inheritance. The data for Ghana discussed in the previous section

showed that women and men in households with lower levels of assets were more likely

than individuals in higher asset quintiles to have experienced physical, psychological and

economic violence. Women in the lower asset quintiles were also less likely to have

experienced social violence than women in higher asset quintiles. However, social violence

among men and sexual violence among both men and women was more concentrated

among those in the top asset quintiles.

The regression analysis in Table E1 (in Appendix E) confirms that the asset ownership

index is negatively and significantly associated with psychological and economic violence

among women: for each of these forms of violence, a one unit increase in the asset index is

associated with a five per cent reduction in the odds of a given woman experiencing

domestic violence. The results in Table E1 indicate that an increase of one standard

deviation of the index is associated with a reduction in psychological and economic

violence among women of almost a third (32.5 per cent) – a very large effect. Asset

ownership is not statistically associated with any other form of domestic violence against

women – indicating that low levels of wealth are only significant risk factors for

psychological and economic violence among women.

The results among the male sample are different. Asset ownership is not statistically

associated with the odds of men experiencing psychological, physical or economic

violence. However, asset ownership is positively correlated with the odds of men

experiencing sexual or social violence (as discussed in subsection 4.1.2): the likelihood of

men experiencing sexual or social violence increased by 3 and 2 per cent, respectively,

when the asset index increases by one unit. This result confirms that men of higher

economic status were more likely than men in lower asset quintiles to have experienced

social and sexual violence in the 12 months prior to the survey.

Even though higher levels of asset ownership lowered the likelihood of some forms of

domestic violence (particularly among women), the qualitative data showed that one of

the major causes of fights in families who experienced domestic violence were disputes

over land assets. This finding was particularly strong in the northern regions, where

190

gender norms played a major role in land conflicts in polygamous contexts that were

arbitrated by customary law. Tensions were particularly high when, for instance, the first

wife of a deceased man had only girls and the second wife had boys: the family of the

second wife could in principle push for the girls not to inherit any of their deceased father’s

property. If properties were not registered, families would have to divide the property

among themselves, which often raised suspicions and mistrust. A female key informant,

who works in litigation in the Northern region, explained when these conflicts occur:

“When one person tries to cheat other family members in sharing family property or when

one person tries to dispose of a family property without the approval and consent of other

family members or to the detriment of the other family members.”

In addition:

“If the man has more than two wives when he dies, the sharing of his property becomes a

problem. Because everybody wants to inherit the man’s property and that brings

disintegration into the family.” (FGD, Upper West region, Jipara district: female group,

18–50 years)

Focus group discussions across all ten regions also referred to fights over assets between

siblings, and within the wider kinship network, when parents died as determinants of

domestic violence. The following excerpt from an interview with an opinion leader in the

Ashanti region further captures some of the dynamics of conflict among siblings that were

frequently cited as causing violence across households in the region:

“Land matter is one of the disagreements and fights in families. The sharing of the land for

farming purposes becomes a problem.”

The same person added:

“There is fighting in the family because of cheating in the sharing of the family properties

like houses, lands, cattle, sheep, goats or trying to take what rightfully belongs to the other

family member(s).”

Education. Section 4.2 discussed how women with primary education were more likely

than women with other levels of education to have experienced physical, psychological or

economic violence. Social and sexual violence was more predominant among women with

secondary education. There was a much less clear pattern of association between education

and domestic violence among men, but men with higher levels education were more at

risk of social and sexual violence.

The multivariate regressions show very varied patterns of association between education

levels and domestic violence among women: (i) women with primary school education are

82 per cent more likely to have experienced physical domestic violence than women with

no education; (ii) women with secondary education are 2.5 times more likely to have

experienced sexual violence than women with no education; (iii) women with primary,

middle and secondary education are, respectively, 82, 89 and 71 per cent more likely to

have experienced social violence than women with no education; and (iv) women with

191

primary education are 50 per cent more likely to have experienced economic violence than

women with no education. There was no statistically significant association between

education levels and the incidence of psychological domestic violence among women

(Table E1, in Appendix E). In general, however, it appears that women with no education

are less likely to experience any form of domestic violence. These results may, however, be

due to the fact that women with more education are also more likely to report having

experienced domestic violence, a factor highlighted in the international literature and in

previous studies in Ghana, as discussed in Section 2.

Education is largely unrelated (or only weakly statistically associated) to the likelihood of

men experiencing domestic violence. The exceptions are economic and sexual violence.

Men with primary, middle, secondary, technical or higher education are, respectively, 2.8,

2.9, 2.9, 3.7 and 2.5 times more likely to experience economic violence than men with no

education. Men with no education seem to be at lower risk of experiencing this form of

violence. Education also matters significantly for the incidence of sexual violence among

men: men with any level of education are substantially more likely to experience sexual

violence than men with no education, for whom the likelihood of experiencing sexual

violence is very close to zero (Table E2, in Appendix E).

Poverty and non-compliance with economic responsibilities. Several respondents in the

focus group discussions and in-depth interviews identified the failure of men to provide

for their family as a major cause of conflict within families, as discussed in the previous

section. Verbal insults and public humiliation by women often followed the failure of men

to provide adequate economic resources, and, in turn, were seen as triggering the use of

physical violence by men against women. This view was widely held across all regions and

most discussion groups, as exemplified in the following statement:

“When a woman nags or complains a lot if her husband is unable to provide her needs, it

obstructs the man from having peace of mind at home. And this makes him go out of the

house often and returns late in the night.” (FGD, Upper East region, Garu-Tenpane

district: female group, 18–50 years)

The GFLHS 2015 data confirmed similar dynamics: 9.4 per cent of men who never denied

chop money suffered from psychological domestic violence; this rate increased to 19.9 per

cent for male respondents who ever denied chop money.

Across all ten regions, the respondents – in particular, older married women who

participated in the focus groups – made links between these forms of domestic violence

and the overall structural economic situation beyond the household, which has left men

(and women) unable to provide for their households. These structural constraints included

lack of education and business skills, as well as the availability of few paid jobs in the wider

economy. In a context characterised by few formal employment opportunities, many men

and women were often compelled to set up their own business, many without having

acquired adequate skills. Frustration and the inability to cope with these wider structural

economic problems were identified as factors that fuelled domestic tensions and resulted

192

in psychological violence committed by women against men, and in physical violence

committed by men against women.

The quantitative data allowed some of these findings to be explored in detail. In particular,

the research team analysed the association between employment status and domestic

violence, because employment opportunities are typically a good indicator of the overall

wealth of the economy, and a determinant of households’ economic welfare. We discussed

in section 4.2 how, overall, women and men who were not working at the time of the

survey were more likely to experience any type of domestic violence than women and men

who were employed or self-employed. The exception was sexual violence – as discussed

before, this form of violence was more prominent among women and men who were

employed.

However, the regression analysis in Tables E1 and E2 (both in Appendix E) did not support

these results. These tables show that there is almost no correlation between the

employment status of women or men and their experiences of domestic violence,

suggesting that other factors associated with employment status may be driving the results

(for instance, asset ownership or education). The main exception is domestic economic

violence: (i) women who were not working were 43 per cent more likely to have

experienced economic violence than self-employed women; and (ii) men who were

employed or not working were, respectively, 2 times and 2.5 times more likely to have

experienced economic violence than self-employed men. In line with previous discussions,

employed men are at greater risk of sexual violence (Table E1, in Appendix E).

Timing of economic events and expenditures. Related to the findings above, domestic

violence was also found to be linked to specific moments, festivals, events and seasons,

such as Christmas, the beginning of the school year and pay day. An opinion leader in the

Eastern region summarised this issue, widely reported across all regions:

“Yes, normally some of these fights happen during funeral times, festivals and also when

schools resume and parents have to pay school fees, sew new school uniform and other

stationery.”

In addition, conflicts about expenses at the start of school were reported across Ghana.

Respondents from all regions mentioned that women usually buy the teaching and

learning materials for their children with financial support from their husband. When

children are not properly clothed or are missing items, they might be sent home, bringing

shame to their family. Fighting about school expenses predictably takes place, as illustrated

in this quote, which reflects nationally representative findings:

“When schools reopen, usually within the second and third weeks. It is during this period

that we need the children’s learning materials. So this information is put on paper to be given

to their parents, and this can cause fights if fathers refuse to follow that.” (IDI, opinion

leader, Northern region, Sagnerigu district)

193

In farming communities, harvesting or planting seasons were often mentioned as a cause

of conflicts between couples, particularly when views diverged as to whether to send

children to school or keep them at home to help on the farm. The following respondent

expresses a view often encountered in rural areas:

“Mostly in the rural and farming communities, it happens between the months of May to

October, when the first rains fall and the farmers plant their crops. Often the man may ask

the children to stay home from school and help with the farm work, the woman may object to

this and it often results in violence.” (IDI, opinion leader, Northern region, Sagnerigu

district)

Spending monthly salary payments on personal pleasure – especially on drinking – was

found to lead to domestic conflicts. A community leader articulated this nationally

representative finding, saying:

“When workers are paid at the end of the month and the man does not cater for the family,

but is seen at drinking spots, it can also cause fights within the families.” (IDI, Eastern

region, East Akim district)

Alcoholism was mentioned in every region, and almost every district, as a particularly

problematic determinant of domestic violence, with men who consume excessive levels of

alcohol being more likely to commit violence against their wives. This was more frequently

articulated to be a problem in communities where pito or akpeteshi (local beers/spirits)

consumption was high in the poorer northern regions. For example, a community leader

explained that:

“Some of the men force the women to go and work on farm, whiles they sit at home,

sometimes drinking ‘pito’ – that is the local alcohol drink.” (IDI, Upper East region, Garu

Tempane district)

The regression analysis in Table E1 (Appendix E) showed that alcohol consumption was

significantly associated with 39 per cent higher odds of women experiencing economic

violence. Alcohol habits are associated with a 66 per cent increase and a doubling of the

odds of men experiencing psychological and sexual violence, respectively (Table E2, in

Appendix E). These results do not indicate a causal relationship between alcohol

consumption and domestic violence but strongly suggest that the risks of domestic

violence incidence increases when alcohol consumption is high.

4.5.3. Marital status

Several international and Ghana-based studies reviewed in Section 2 have shown that

marital status is an important factor associated with domestic violence. The discussion in

Section 4.2 on incidence of domestic violence showed mixed patterns: married women (or

those living with a partner) were at higher risk of sexual or economic domestic violence

than unmarried women, whereas divorced, separated or widowed men were more at risk

of social or sexual violence than married or never-married men. All other forms of violence

were more prominent among non-married women and men.

194

These results change substantially in the regression analysis in Tables E1 and E2 (Appendix

E). The results show that, in general, not being married reduces the likelihood of having

experienced domestic violence among women and men. Furthermore:

Divorced, separated or widowed women were twice as likely to have experienced

physical violence than women who were never married.

Married women or divorced, separated or widowed women were, respectively, 1.8

times and 2 times more likely to have experienced economic violence than non-

married women.

Married or divorced, separated or widowed men were, respectively, 6 times and

almost 8 times more likely to have experienced sexual violence than non-married

men, and almost 3 times and 1.7 times more likely to have experienced social

violence than men who were never married.

Table E1 (Appendix E) shows that women in polygamous relationships were almost twice

as likely to experience psychological violence than women who were never married. These

women were also 3.5 times more likely to have experienced physical violence and 4.2 times

more likely to have experienced economic violence. Among men, polygamy increases

considerably the odds of sexual or social violence against them (Table E2, in Appendix E).

However, the association between the respondent’s marital status and experiences of

different types of domestic violence could reflect reverse causality, as divorce and

separation, for instance, may result from domestic violence. Disentangling these issues

through quantitative data alone is very challenging. The qualitative data (and some

additional quantitative data) suggest some plausible mechanisms that may explain these

correlations, including access to sex within marital relations, extra-marital affairs and

forced marriage, as discussed in more detail below.

Access to sex within marital relations. The work of Ellsberg et al. (2001), cited in Section

2, showed how, in many parts of the world, marriage is seen as guaranteeing men

unconditional sexual access to their wives. Conflict over sex was a ubiquitous finding in

the qualitative data, prevalent across all regions, and in both rural and urban areas. The

different compositions of the focus groups enabled frank discussions about these sex-

related conflicts, particularly in the single-sex groups of married women and of married

men (aged 18–50, and aged 50+). The principle source of conflict, across all regions, centred

on men’s desire for sex at times when they were unable, or unwilling, to provide for the

members of the household, including giving their wives chop money. This nationally

representative finding was highlighted as one of the major causes of violence – particularly

psychological violence and controlling behaviour.

This finding suggests a close link between economic and marital determinants of domestic

violence: women described how they expressed anger against their partner by refusing sex,

while men responded or initiated marital conflicts by refusing to fulfil their economic

obligations. In particular, most respondents across all regions agreed that both women and

195

men use sex as a weapon that has many forms, including being forced to have sex or being

denied sex. Sex is part of an intricate power play between men and women, as explained

in this interview with an education professional:

“Fear can also make women feel [that] they need to accept the dominant or violent ways of

the men. She is afraid to be beaten, afraid to be denied food and afraid of being driven out of

the matrimonial home, and some even fear being denied sex, especially in the polygamous

marriage. The man can proudly say ‘I will not have sex with you.’” (IDI, Northern region,

Sagnarigu district)

The quantitative survey included a number of questions about this important issue.

Opinions diverged widely about whether women have a right to refuse sex. For instance,

8.2 per cent of respondents (9.8 per cent of women and 5.7 per cent of men) felt that the

refusal of sex was a justifiable reason for wife-beating (Table 55). Employed respondents

(4.6 per cent) were less likely to agree that the denial of sex was a reason for wife-beating

than self-employed (9.2 per cent) and non-working respondents (8.1 per cent); as were

respondents living in urban areas (4.8 per cent, compared to 11.7 per cent in rural areas),

in the Greater Accra region (1.7 per cent, compared to 24.4 per cent in the Northern region,

24.0 per cent in the Upper East region and 29.4 per cent in the Upper West region), with

higher levels of education (2.5 per cent, compared to 17.5 per cent among those with no

education) and in the highest asset quintile (3.1 per cent, compared to 13.4 per cent in the

lowest quintile).

Extra-marital affairs. Men and women who engaged in extra-marital affairs were

perceived to be at risk of domestic violence. The qualitative data indicated that this

violence differed by sex, but that it was ubiquitous across all regions. Women who had

extra-marital affairs were more likely to experience physical violence perpetrated by their

partners, whereas men were more likely to experience psychological violence and

controlling behaviour from their partners. There were, however, some striking anomalies

in the qualitative data, which suggested that this gendered dichotomy of experiences of

violence might not always hold true. In a focus group with young men and women (aged

18–25) in the Greater Accra region, this young person recalled a case of a man who found

out that his wife was cheating:

“What happened was that the man was called by the friend that he saw the wife somewhere

with another man. Instead of the man asking the woman where she went, upon her return,

he pounced on her and started beating her and insulted her. The woman also went into the

room and got herself an iron and hit the man with it.” (Mixed FGD, Gbawe,

unmarried/youth (18–25 years)

The findings from the survey differed somewhat from the qualitative findings but are

equally paradoxical: 0.4 per cent of the survey respondents agreed that unfaithfulness

would be an acceptable reason to hit women, but 7.5 per cent mentioned that the suspicion

of unfaithfulness was an acceptable reason for wife-beating (Table 55). This finding

requires further future research, as the data collected are not sufficient to fully understand

these findings: unfaithfulness emerges as a risk factor for domestic violence, but

196

perceptions and responses vary substantially across types and perpetrators of domestic

violence.

Forced marriage. Forced marriage may itself be considered a form of domestic violence

perpetrated by families against girls with the consent of their parents. While not widely

reported in the qualitative data, respondents in the Northern region reported that parents

would be more likely to arrange an early or forced marriage when girls were found to be

underperforming in school and paying attention to boys. While non-representative, this

narrative was iterated by a number of community leaders and activists in interviews. For

example, according to a faith leader in the Northern region:

“As we sit now, there is a girl whose parents have arranged with an old man who is already

married to two wives to support their daughter in school and marry her. The girl is now in

senior high (SHS) and this is creating a serious fight between the girl and her parents.” (IDI,

Northern region, Sagnarigu district)

Similarly, an activist working on education explained that early marriages are a form of

violence that reflect structural poverty and absent fathers:

“Some of the children who are female may look for support from outside the house from men.

No man will give out money free. Therefore, these girls may end up being impregnated. This

is where early marriage comes in. She may be forced to be given out to marriage at that tender

age, if she attends school, she may drop out of school making her situation worse and her

future bleak.” (IDI, Northern region, Sagnerigu district)

This issue was only raised in the discussion groups in the northern regions of Ghana.

4.5.4. Geographical effects

The results showed several differences in how domestic violence and attitudes towards

domestic violence were reported across the different regions of Ghana. The discussion in

Section 4.2 highlighted that women and men living in urban areas were at greater risk of

experiencing social or sexual violence, whereas women and men living in rural areas were

at greater risk of experiencing physical, psychological or economic domestic violence. The

experiences of women and men across the ten regions were very different, and no

immediate pattern stood out.

The regression analysis in Tables E1 and E2 (Appendix E), which control for a range of risk

factors, showed no statistically significant difference in the likelihood of domestic violence

experienced by women in urban and rural areas (the coefficients were not statistically

significant). The only exceptions are that: (i) women in rural areas were 39 per cent more

likely to suffer from economic violence than women in urban areas; and (ii) men in rural

areas were 2.2 times more likely to have experienced physical violence than men living in

urban areas. In all other cases, the effect of living in a rural or in an urban location was not

found to be associated with domestic violence.

197

Most of the regional differences observed in women’s experiences of physical and social

violence disappear in the multivariate analysis (Tables E1 and E2 in Appendix E). The only

associations found between regions and the incidence of domestic violence included: (i)

psychological violence against women in the Central, Eastern, Ashanti and Upper regions,

and against men in the Eastern, Brong Ahafo and Upper East regions; (ii) physical violence

against women in the Eastern and Upper East regions; (iii) sexual violence against women

in the Central and Greater Accra regions; (iv) social violence against women in the Eastern

region and against men in the Brong Ahafo, Northern and Upper East regions; and (v)

economic violence against women in the Central, Greater Accra, Volta, Ashanti, Northern,

Upper East and Upper West regions, and against men in the Volta and Upper East regions.

Table 60 summarises this information.

Table 60: Regions significantly associated with increased experience of domestic

violence (by type of domestic violence and sex)

Social

violence

Physical

violence

Sexual

violence

Psychological

violence

Economic

violence

Women Eastern Eastern

Upper East

Central

Greater

Accra

Central

Eastern

Ashanti

Upper East

Upper West

Central

Greater Accra

Volta

Ashanti

Northern

Upper East

Upper West

Men Brong Ahafo

Northern

Upper East

Eastern

Brong Ahafo

Upper East

Volta

Upper East

Source: Ghana Family Life and Health Survey (GFLHS) 2015 and DHS 2008.

Note: Based on results reported in Tables E1 and E2 (Appendix E).

However, for both the male and female samples, regional effects were not collectively

statistically significant (Tables E1 and E2, in Appendix E).31 These results suggest that, once

the regressions control for a range of variables, there is no regional exceptionalism in

patterns of domestic violence. In other words, the differences in domestic violence

incidence discussed in Section 4.2 are due to a range of factors – possibly levels of poverty

or economic development – but not to intrinsic geographical characteristics.

4.5.5. Patriarchal and gender norms

Section 2 discussed in detail how patriarchal norms and perceptions of masculinity that

support the persistence of certain gender norms may be closely associated with experiences

of domestic violence. The association of domestic violence in Ghana with patriarchal

31 With the exception of economic violence, in the regression for men, for which regional effects are

marginally significant.

198

cultural norms was evident in the qualitative and quantitative data collected across all

regions and cut across the various socio-economic backgrounds of the respondents.

The qualitative and quantitative data found that social norms concerning appropriate

behaviour for women were a significant trigger for domestic violence by men against

women in all regions. In the qualitative interviews and focus groups, the research team

asked which expectations of masculinity make men behave in certain dominant or violent

ways. The male and female respondents, in every region, discussed in detail which societal

gender norms and expectations shaped appropriate and inappropriate behaviour for both

men and women. The dominant view was that domestic violence was associated with non-

compliance with these gender norms. As discussed in the section above on attitudes to

violence, the quantitative and qualitative data found that different forms of domestic

violence were used to ‘punish’ men and women, and young people, when they had been

found to contravene normative expectations of appropriate behaviour. For instance, the

qualitative data found that men were more likely to become violent when their expectation

of being obeyed was not fulfilled by their wives, or when wives failed to perform expected

domestic roles, such as ensuring their meals were ready on time, or refused to have sex. A

community elder in the Ashanti region explained a widely held view across all regions:

“Men, like children, expect to be pampered at home. Thus, if they do not get this from their

wives, it makes them act violently in the house. Also when they are refused sex it makes them

go wild in the house. The other main reason why men act violently at home is when women

do not [fulfil] their wifely duties. For instance, you know the man sometimes come back from

town in the company of friends and when the wife always leave the house untidy, he will

fight with you.”

The respondent added:

“The man expects that he would be accord[ed] the needed respect and dignity at home by

both his wife and children so, if he is not getting it, it becomes a problem. It also happens

when the man sees that the woman is inciting the children or trying to organise the children

against him. The man also expects that his instructions must be obeyed at home.”

The qualitative data from all regions, and especially from the Northern region, also found

that societal norms governing expectations about women made them more accepting of

abusive relationships. Both male and female respondents pointed to the ‘expected’

submissiveness of women within conjugal relationships, women’s roles as primary

caregivers (for children) and their economic dependence on men as reasons for them to

accept violence. For example, a health practitioner in the Ashanti region discussed these

gender dynamics in an interview, saying:

“At times they can say I am the head of the house. Whatever I instruct you to do, you do it.

I give you money daily for you to eat. I clothe you. That is why at times they do that. It could

also be our bringing up as Ghanaians and also what they tell married couples on the day of

their marriage. Whatever your husband tells you to do, do it. Even if your husband beats

you, stay in the marriage. The marriage counselling she received says that the husband is the

head. Even in our churches they say the husband is the head. Maybe it could be that the

199

woman does not wash the man’s clothes, or not giving the man food to eat. When you also

deny the man sex, he can behave in this violent way.”

This view emerged across the focus group discussions in all ten regions, and is illustrated

by the following exchange in the Central region:

Respondent 8: We are told from the word go that we have to be submissive so we seem to

have no other choice. Sometimes is because of the children so we don’t even care what happens

to us. I will plead with men to have patience for women.

Respondent 7: Sometimes is because the man takes care of all finances so we can only be

submissive no matter what the case may be but in cases where at least the woman is engaged

in some occupation and can also support financially then the domination will be minimised.

Respondent 2: The men take care of us, so we cannot even go anywhere without their

permission.

(FGD, Central region, Anomabo district: male and female group, 18–25 years)

The quantitative data allowed the research team to analyse the effect of gender norms on

domestic violence using four distinct variables: a patriarchal gender norms index; an index

that measured tolerance towards wife-beating; an index that measured acceptance of

women’s sexual autonomy; and a measure of decision-making power within the

household.

The results in Table E1 (Appendix E) show that women who exhibited higher levels of

acceptance of wife-beating were 14 per cent more likely to have experienced economic

violence than women with lower acceptance of wife-beating. As mentioned before, this is

not a causal effect and may well suggest (as discussed in Section 4.2 on attitudes) that

victims of violence internalise and accept gender norms that perpetuate or justify domestic

violence. Women who expressed a higher acceptance of women’s sexual autonomy were

16 per cent more likely to have experienced psychological and social violence, 17 per cent

more likely to have experienced physical violence, and 11 per cent more likely to have

experienced economic violence than women with a lower acceptance of women’s sexual

autonomy. This latter finding suggests, as discussed in the qualitative data analysis above,

that women who are more likely to challenge accepted gender norms may be at risk of

domestic violence because they may be punished for their non-compliance with

established norms of behaviour. The results in Table E1 (Appendix E) show no statistically

significant association between domestic violence against women and the patriarchal

gender norms index or the levels of decision-making within the household.

The results in Table E2 (Appendix E) show that men who expressed adopting stronger

patriarchal gender norms were 26 per cent less likely to report social violence than men

who did not accept such norms. However, men who displayed higher levels of tolerance

towards wife-beating were 21, 47 and 30 per cent more likely to have experienced

psychological, physical or economic violence, respectively, than men who reported lower

levels of tolerance towards wife-beating. This result may potentially reflect retaliation by

women who are beaten by their husbands, as discussed in the previous section. Men who

200

exercised more decision-making power within their households were 34 per cent less likely

to experience social violence than men who exercised less levels of decision-making within

the household.

Taken together, and in conjunction with the qualitative data discussed above, these results

suggest that gender norms shape the extent of domestic violence in Ghana. This finding is

in line with recent research by Heise and Manji (2016), who have shown how social norms

play important, albeit different roles, for actors functioning at different levels of the social

ecological model.

4.5.6. Exposure to other forms of violence

Some recent evidence from conflict-affected countries has suggested that domestic

violence may be associated with other forms of violence that take place across countries

and within communities. The survey questionnaire included questions on various forms

of violence that may occur in the Ghanaian society, which were aggregated using an index

that included how many of the following types of violence the respondent had witnessed

in their community: theft or robbery, vandalism, arson, kidnapping or abduction,

extortion, bullying, political violence, highway attacks, harassment, rape, defilement and

seduction. The regressions in Tables E1 and E2 (Appendix E) show that exposure to

violence in the community was significantly associated with higher levels of domestic

violence. The odds of victimisation among women increased by 77 per cent for

psychological violence, 49 per cent for physical violence and 57 per cent for economic

violence when the respondent lived in a community that experienced at least one form of

violence. The effect was stronger for sexual violence and social violence, for which the odds

of having experienced this violence were 3.6 times and 1.9 times higher, respectively, for

women exposed to violence in their communities than for women in communities that did

not experience other forms of violence.

The results showed an even stronger association between violence in the community and

incidence of domestic violence among men. Men who reported some violence in their

communities were between two and three times more likely than men living in more

peaceful communities to report having experienced any form of domestic violence.

The qualitative data emphasised, in addition, that in the northern regions of Ghana

chieftaincy disputes between extended royal family members can erupt into community-

wide violence if, for instance, the members of the extended family or even siblings fight

over the right to the throne, or disputes and fights erupt between other family members

who may support different candidates:

“Sometimes too if the man and woman’s family are fighting over the right to the throne, it

often results in fights and conflicts between them and sometimes a break of the union by the

two families even when they love each other.” (IDI, Northern region, Sagnerigu district,

Gbolo opinion leader)

201

Chieftaincy disputes were only mentioned in the northern regions, where they are

described as “a major cause of disagreements and fights in this community.” (FGD, Upper East

region, Garu-Tempane district: male group, 18–50 years)

Despite the strong association between domestic violence and other forms of community

violence, the study is not able to indicate the direction of causation. On the one hand,

communities that experience high levels of community violence may also experience high

levels of domestic violence because violence becomes an accepted way of behaviour. On

the other hand, it is possible that experiencing violence within the family and immediate

social relations may predispose individuals to other forms of violence. This result merits

further research in future analyses of the GFLHS 2015.

4.6. Consequences of domestic violence in Ghana

This section addresses the research question about the consequences of domestic violence

in Ghana. The main consequences of domestic violence emphasised in the literature

reviewed in Section 2 included effects on physical health, effects on mental health and

emotional well-being, effects on concentration during and take-up of day-to-day tasks, and

indirect effects on children. The qualitative and quantitative data collected in Ghana for

this study revealed similar patterns, with a strong emphasis on the negative physical and

mental health impacts of domestic violence. As with other sections, this analysis draws on

complementarities between these different data. Multivariate regression analyses were

also undertaken to explore the physical and mental health consequences of domestic

violence. The study used three self-reported measures of physical health: health status,

illness and chronic condition. Health status was captured by the following question: ‘In

general, would you describe your general health as very good, good, fair, poor or very

poor?’ ‘Very good’ and ‘good’ categories were grouped together into a binary variable

indicating ‘good health’. Illness was captured by the following question: ‘In the last 30 days,

have you felt sick or unwell?’ Chronic condition was measured by the following question:

‘Do you have any chronic condition?’.

The study also used the K6 scale of mental illness developed by Kessler et al. (2003) to

assess mental health. The K6 scale is constructed from a survey module that asked how often

in the past 30 days the respondent felt any of these six symptoms: nervous, hopeless,

restless, so depressed that nothing could cheer them up, that everything was an effort, and

worthless. For each of these six questions (K6), possible answers included: ‘all the time’,

‘most of time’, ‘some of the time’, ‘a little of the time’ and ‘none of the time’. A respondent

who answered ‘all the time’ across the six items was attributed a maximum K6 score of 24

(6*4). Respondents with K6 scores exceeding 13 were considered to be suffering from

serious mental illness (SMI) (Kessler et al., 2010). The regression analysis used both the

continuous (K6 scale) and the binary (SMI) variables in the analysis. Regressions, in

addition, control for age, education, marital status, socio-economic status, residence and

violence in the community, as well as food security (which takes the value 1 if household

members went to bed hungry, even if only rarely, during the last lean season), source of

202

drinking water (distinguishing between piped water into the dwelling; public taps, wells

and springs; surface water; and bottle/sachet water) and access to toilet (distinguishing

between private toilet; shared toilet between a few families; public toilet; and no toilet).

4.6.1. Physical health effects

To measure the impact of domestic violence on respondents’ health, the quantitative

survey asked respondents who reported having experienced any act of sexual or physical

violence whether the act had resulted in injuries. A total of 182 men and 315 women across

the whole GFLHS 2015 experienced an act of sexual violence in the 12 months preceding

the survey, of which 32 incidents were domestic acts of violence against men, and 78

against women. Out of these, nine women and one man reported having received injuries.

Seven women were injured once or twice, and the man and two women more than five

times. Four of the respondents needed health care following their injuries; one of the

women needed health care three times in the 12 months prior to the survey.

Fifty men and 192 women survey respondents reported having experienced physical

domestic violence in the 12 months that preceded the survey. Injuries were reported by 114

of these respondents: 14 men and 77 women were injured once or twice during that time

frame, one man and 16 women were injured between three and five times, and one man

and five women were injured more than five times. Forty-six women and ten men were

hurt badly enough to have needed health care: 32 once, and 24 more often.

Other studies conducted in Ghana (e.g. Pool et al., 2014) have shown that violence against

women is associated with serious health consequences in terms of pregnancy outcomes.

The GFLHS 2015 found that 30 of the 192 women who reported having been physically

assaulted were hit while pregnant (15.6 per cent), resulting in a miscarriage rate of 10 per

cent (three women) and neurological complications to the foetus in 6.7 per cent (two

women) of the cases. Another 6.7 per cent reported other, non-specified, effects on the child

or mother.

Domestic violence also appears to have exerted an adverse impact on general health

outcomes: 43.8 per cent of women who had been physically assaulted declared having been

ill in the 30 days prior to the survey – a rate that was one third higher than among women

who were not physically assaulted (31.2 per cent). Also, 42.3 per cent of women who had

experienced psychological violence reported ill health, compared to 30.9 per cent of

women who did not experience this form of domestic violence.

The social ecological model adopted by this study understands the consequences of

domestic violence within the complex interplay of individual, relationship, social, cultural

and environmental factors. This understanding was evidenced in all focus group

discussions and in-depth interviews that addressed the consequences of domestic violence,

and particularly emphasised in the younger (mixed unmarried, 18–25 years) and older

groups (mixed, 50+ years). These two groups mentioned, in addition, that the health

203

consequences of domestic violence were felt not just by the individuals involved (victim or

perpetrator) but also by witnesses of the violence within the household, family and

community.

At the individual level, domestic violence was described as affecting several health

complications such as increased blood pressure, mental disorders, depression, eating

disorders, sleeplessness, alcoholism and suicidal thoughts. These responses sum up the

range of views on the health consequences of domestic violence:

“Psychological violence is also likely to cause the victim stress and pressure because

continuously he or she would have to be thinking and feel isolated. These could also lead to

several health complications such as increase in blood pressure, mental disorders, depression,

eating disorders, sleeplessness, alcoholism, suicidal thoughts and others. Sexual violence also

results in unwanted pregnancy, chronic pelvic and abdominal pain and contraction of STIs.”

(IDI, opinion leader, Kukurantumi/Koforidua East Akim district, Eastern region)

Physical violence was commonly noted as affecting the physical health of women and men.

For men in particular, and in all regions, alcoholism was consistently mentioned by all

categories of focus group discussions and in-depth interviews as both a consequence and

a cause of violence (as discussed also in the previous section). This nationally

representative view is well expressed by a respondent in a focus group discussion in the

Greater Accra region:

“It [domestic violence] affects the man and woman mentally, emotionally, psychologically,

to the extent that the men start drinking alcohol and the woman suffer from depression.”

(FGD, mixed 50+, Tomefa, Ga South district, Greater Accra region)

Focus group discussions across all regions mentioned that household and community

members who witnessed these outbursts of violence were also negatively affected by the

violence, as illustrated in these two poignant quotes:

“[The] aged in such environments could have their blood pressures rising upon hearing of a

fight.” (FGD, mixed 50+, Gbawe, Ga South district, Greater Accra region)

“[Domestic violence] causes emotional distress to neighbours or family members who may

be in the household.” (FGD, married/living together women 18–50, Gbolo, Sagnerigu

district, Northern region)

These responses reveal the relationship between individual and contextual factors and

demonstrate how domestic violence affects multiple social spheres.

Table 61 shows the regression results for the effect of each form of domestic violence –

psychological, physical, sexual, social and economic – on the physical and mental health

of women. The results show that:

Women who experienced psychological violence were 29 per cent less likely to

assess their health as very good or good than women not affected by violence.

Women who experienced physical violence were 67 per cent more likely to have

had an illness in the month prior to the survey than women not affected by violence.

204

Women who experienced social violence were 32 per cent less likely to have

assessed their health as very good or good than women not affected by violence.

Women who experienced economic violence were 40 per cent less likely to assess

their health as very good or good, 36 per cent more likely to report having been ill

in the month prior to the survey and 66 per cent more likely to have reported

suffering from at least one chronic condition than women not affected by violence.


(women)

(1) (2) (3)

Health status

(1=very

good/good)

Illness (1=ill

during last

month)

Chronic conditions

(1=yes)

Logit Logit Logit

Psychological violence 0.71* 1.19 1.10

(0.13) (0.19) (0.23)

Physical violence 0.71 1.67*** 1.38

(0.16) (0.29) (0.34)

Sexual violence 0.97 1.30 0.80

(0.34) (0.39) (0.29)

Social violence 0.68** 1.41** 1.20

(0.13) (0.20) (0.21)

Economic violence 0.60*** 1.36** 1.66***

(0.091) (0.19) (0.30)

Observations 2,971 2,971 2,963


Note: The table shows the odds ratios. Standard errors in parentheses are clustered within enumeration areas

(EA). * p < 0.1, ** p < 0.05, *** p < 0.01. Each column represents the results of a regression equation that includes

all controls included in the tables in Appendix E.

The results in Table 61 show that sexual violence among women did not have any

statistically significant effects on their health. In stark contrast with the results for women,

Table 62 shows that none of the forms of domestic violence were statistically related to the

physical health status of men.

205


(men)

(1) (2) (3)

Health status

(1=very

good/good)

Illness (1=ill

during last

month)

Chronic conditions

(1=yes)

Psychological violence 0.54 0.76 0.94

(0.26) (0.29) (0.46)

Physical violence 0.67 1.55 0.99

(0.71) (0.80) (0.69)

Sexual violence 0.57 2.23

(0.48) (1.67)

Social violence 0.43 1.00 1.63

(0.26) (0.33) (0.78)

Economic violence 0.55 1.37 1.32

(0.25) (0.41) (0.64)

Observations 695 695 686



(EA). * p < 0.1, ** p < 0.05, *** p < 0.01. Each column represents the results of a regression equation that includes

all controls included in the tables in Appendix E.

The higher incidence of health effects of domestic violence among women may be due to

women being more vulnerable to physical assault than men. As a respondent poignantly

pointed out:

“The abusers are natural[ly] stronger. So if it comes to the physical attack very few women

will be able to outwit men. Most times the men overpower the women.” (IDI, opinion

leader, Jirrapa district, urban Jirrapa)

As discussed in Section 4.1, this view was common across all regions, as illustrated by this

quote:

“There is a difference in the violence that men and women experience because a man is

stronger than a woman. As a result, women are mostly hurt in violence than men.” (FGD,

Upper West Jirrapa district, rural Wuling)

And, when

“…the violence is physical, it can cause injury to the victim which could lead to permanent

deformity in the long term.” (FGD, married/living together men, Gbolo: Sagnerigu

district, Northern region)

206

The qualitative data showed that the consequences of physical abuse, including slapping,

pushing, kicking, strangling and using weapons, varied from minor injuries to death.

Physical injuries often led to hospitalisation and incapacitated victims:

“Women experience a lot of domestic violent cases more than men. A lot of women end up

losing their lives because they are most often at the receiving end of these brutish attacks.”

(IDI, opinion leader, Goaso, Ahafo North district, Brong Ahafo region)

The majority of respondents interviewed who work in health services mentioned having

seen permanent disability, scars and even death as a result of physical violence,

particularly against women.

4.6.2. Mental health and emotional effects

The survey data reported a very strong correlation between exposure to domestic violence

and mental health. Using the K6 scale for depression and anxiety, Table 63 shows that the

likelihood of suffering from serious mental illness (i.e. having K6 scores exceeding 13) was

considerably higher for women and men exposed to domestic violence in the 12 months

prior to the survey than for those who were not. Women who reported serious mental

illness were 26.7, 24.2, 21.9, 19.3 and 26.3 per cent more likely to have experienced,

respectively, psychological, physical, sexual, social or economic violence than women who

did not report serious mental illness. The association between serious mental illness and

domestic violence was generally lower among men, but still significant across all

categories, with the exception of sexual violence.

Table 63: Percentage of respondents who have experienced domestic violence

reporting serious mental illness (by type of domestic violence and sex)

Psychological

violence

Physical

violence

Sexual

violence

Social

violence

Economic

violence

Women

Yes 26.7*** 24.2*** 21.9*** 19.3*** 26.3***

No 11.0 11.7 12.3 11.6 10.6

Men

Yes 11.7* 19.0*** 6.9 12.7** 15.7***

No 7.9 7.9 8.2 7.8 7.6


Note: Pearson Chi-squared test of equality of means between ‘yes’ and ‘no’ categories: *** p<0.01; ** p<0.05;

*p<0.1

Table 64 illustrates further, using regression analysis, the consequences of domestic

violence on mental health among women. Column 3 shows the effect of each form of

domestic violence on their K6 score (women exceeding a K6 score of 13 were considered

to be suffering from serious mental illness). Column 4 shows the same results using a

207

binary variable (which takes the value 1 if the woman was considered to be suffering from

serious mental illness, and 0 otherwise). The results were as follows:

Women who experienced psychological violence had a K6 score that was 5.6

percentage points higher than women who did not experience psychological

violence, and were twice as likely to suffer from serious mental illness.

Women who experienced social violence had a K6 score that was 2.4 percentage

points higher than women who did not experience social violence, and were almost

1.5 times more likely to suffer from serious mental illness.

Women who experienced economic violence had a K6 score that was almost 10

percentage points higher than women who did not experience economic violence,

and were 94 per cent more likely to suffer from serious mental illness.

Table 64: Effect of domestic violence on men and women’s mental health

(1) (2) (3) (4)

Men: K6

score

Men: Serious

mental illness

(K6 > 13)

Women: K6

score

Women:

Serious

mental illness

(K6 > 13)

OLS Logit OLS Logit

Psychological violence 1.22 0.35 5.63*** 2.07***

(0.73) (0.23) (2.58) (0.43)

Physical violence 4.12 5.86*** 2.07 1.18

(5.39) (3.75) (1.10) (0.27)

Sexual violence 0.77 n.a. 2.95 0.94

(1.26) (2.45) (0.38)

Social violence 1.04 0.80 2.41*** 1.47**

(0.75) (0.50) (0.81) (0.24)

Economic violence 3.05 1.64 9.95*** 1.94***

(2.09) (0.90) (4.44) (0.34)

Controls YES YES YES YES

Observations 704 687 2,972 2,972

R-squared 0.12 0.16



(EA). * p < 0.1; ** p < 0.05; *** p < 0.01.

The results show that physical and sexual violence had no statistically significant effects

on the mental health status of women. In addition, Columns 1 and 2 of Table 64 show that

(as with physical health) domestic violence did not have statistically significant effects on

the mental health status of men. The only exception was for physical violence, where those

who have experienced physical violence were 5.8 times more likely than men who did not

experience physical violence to suffer from serious mental illness. Some results from the

qualitative analysis may explain this latter finding. Although this view was not nationally

208

representative, an opinion leader in the Volta region mentioned that some men took to

drinking because of violence at home, which affected their mental capacity to work on the

farm or at any workplace.

4.6.3. Effects on work and daily activities

In line with the discussion in Section 2, the survey included a number of questions about

how each act of physical or sexual violence experienced in the 12 months prior to the

survey affected the daily lives of those who had experienced violence, in terms of their

absence from or ability to work, go to school or do domestic work, their ability to

concentrate on daily activities, their levels of confidence and on reporting feelings of living

in fear.

Thirty-two men and 78 women across the whole GFLHS 2015 reported consequences of

sexual violence. The majority of respondents affected by domestic sexual violence (90.6 per

cent of men and 87.2 per cent of women) reported no impact of sexual violence on their

daily activities. Ten women and four men felt that domestic violence had adversely

affected their ability to concentrate on work. Others (less than three individuals) reported

that they could not focus at school, were unable to concentrate on their domestic chores,

lost confidence in their own abilities and lived in constant fear.

The GFLHS 2015 also asked about the effects of physical violence on the daily lives of

individuals who experienced it. Fifty men and 192 women across the whole survey

reported effects of physical violence. The most prominent consequence for men was

inability to concentrate at work (four men). Among women, 39 were unable to concentrate

while doing their domestic work, 37 were unable to concentrate while at work, 30 lost

confidence in their own abilities, 22 lived in constant fear, and 21 missed work.

The qualitative data also showed that women became economically more vulnerable as a

result of domestic violence. This view was widely mentioned across all regions by

participants in both focus group discussions and in-depth interviews, who discussed the

extra burden placed by vengeful perpetrators as they refused to provide housekeeping or

chop money following other acts of violence. Reflecting this view, an opinion leader stated:

“The man will decide not to provide for the family. He will refuse to give chop money and

also stop paying the children’s school fees.” (IDI, opinion leader, Gbi Kpoeta, Hohoe

municipality, Volta region)

As a result, women often came to accept abusive relationships because of their economic

dependence on men. This economic motivation for the acceptance of violence was not

nationally representative but was more visible in highly patriarchal sites and areas with

greater levels of poverty in the northern regions, as illustrated by the following quote:

“The victims, especially the married women, get scared that when they report their husbands

to the police, they will be jailed and who takes care of them and their children.” (FGD,

Married women, Gbolo, Sagnerigu district, Northern region)

209

4.6.4. Consequences for children

Across all regions of Ghana, respondents in both focus group discussions and in-depth

interviews mentioned the wide-ranging consequences of domestic violence for children

and young people in the short and long term. A prominent effect discussed by respondents

was the psychological impact of domestic violence on children who are deprived of

parental care and socio-economic rights. This nationally representative view was well

illustrated by this quotation:

“Lack of peace in the family can cause the minds of children to retard. Their confidence level

is also reduced as they grow in domestic violence homes.” (IDI, opinion leader, Gbi

Kpoeta, Hohoe district, Volta region)

Respondent after respondent emphasised how care for, and attention to, children may be

compromised in homes that experience domestic violence. A respondent pointed to his

own experience, in a view shared by many in focus group discussions and in-depth

interviews across Ghana:

“It will affect the children emotionally when their daddy is groaning and their mum is

moody. I want to use my experience as an example. I had a little quarrel with my wife, the

child was crying, I was expecting my wife to go for the child and my wife was also expecting

me to go for the child, and because we were angry with each other, we didn’t know who

should go for the child. I just had to be a man and go for the child because she was not ready

to attend to the child.” (IDI, opinion leader, Garu, Garu Tempane district, Upper East

region)

Another way in which domestic violence may affect children was through the loss of their

mother. This effect was only mentioned in areas dominated by patrilineal culture, where

men retain possession of children upon separation, such as in the Upper East and Upper

West regions. This point was illustrated by an interview with an opinion leader:

“When it persists, the woman will seek separation or divorce and will become free from this

burden. When the woman leaves, the children will not have motherly care because in this

area, the children are considered to belong to the man.”

In areas of matrilineal inheritance in the Obuasi district of the Ashanti region, responses

typically pointed to women assuming responsibility for the care of children upon

separation, with children being deprived of paternal care and the economic support that

comes with it:

“Most often it is the children and the woman that suffer especially when they are left in the

care of an unemployed mother. Some men refuse to give house-keeping money for the use of

the mother and the children.” (IDI, opinion leader, Obuasi, Obuasi municipality,

Ashanti region)

210

Children growing up in abusive homes manifest their traumatic experience in school.

Many respondents from the field of education pointed out this link between abusive homes

and challenges in school:

“Yes. Even just this morning, which is a holiday, a woman drove her child to come to class.

Meanwhile the woman knows well that today is a holiday. To provide the child with just

GHc1 is a problem. So she must find ways and means to escape that burden of providing the

child with money, hence the decision to send her away from home. When the child comes to

school, as her friends are eating, she might be lucky to get something to eat. Even today, one

of my teachers gave a child GHc1 because the child was hungry. The moment she collected

the money, she set off for home. When I enquired where her mother was, she replied she was

at home. I enquired further and was told the parents are divorced and for that matter, when

she was coming to school, she wasn’t given any money. At times we are compelled to sacrifice

to feed some of these pupils, providing token money to these young children. Even we have

declared some of these children ‘HIPC’, because anything you ask, they won’t bring it. We

have about 8 pupils in this HIPC category in this school. At times too some of the teachers

buy dresses for these pupils. These are problems we go through. Thus some parents are not

treating their children well at all. This innocent child can be somebody in the near future.”

(IDI, opinion leader, Goaso, Ahafo-Ano district, Brong Ahafo region)

As a result of these experiences, children’s education is also affected, as they start:

“absenting themselves from school or neglect their studies altogether” (IDI, opinion leader,

Tomefa, Ga South district, Greater Accra region)

In addition, echoing a view mentioned across all ten regions:

“Some of the parents refuse to send their children to school. Some of the parents use the

children to do some kind of jobs, which children cannot do. […]Sometime too, threaten them

with dangerous weapons.” (IDI, opinion leader, Boakyekrom, Ahafo North district,

Brong Ahafo region)

These forms of neglect were discussed in particularly strong terms among focus group

discussions and in-depth interviews in areas with high levels of poverty, as this quote

illustrates:

“Some parents abandon their children to the point where some children go to school on empty

stomachs. Parents neglecting their children are a paramount situation here.” (IDI, opinion

leader, Jirapa, Jirapa district, Upper West region)

Children in turn may seek comfort in friends or resort to risky behaviours. The outcome in

the long term may be waywardness and social deviance among the youth, as most opinion

leaders involved in child protection services mentioned:

“The children are not spared under these conditions. They are also negatively affected. They

tend to look for happiness out of home due to the fighting and sometimes fall into bad

company. Some begin to smoke wee, drink and even steal. The girls also look for boyfriends

to support them because their fathers are not caring for them well. This often results in

211

pregnancies which have bad repercussions on their future or career.“ (IDI, opinion leader,

Gbi Kpoeta, Hohoe municipality, Volta region)

In essence, the results from the qualitative data highlighted how domestic violence

experienced in the home, and the resulting divorce or separation, had both short- and long-

term impacts on children’s educational, health, economic and social well-being. This is an

important finding because, as discussed in subsection 4.3.1, experiences of domestic

violence during childhood may themselves be reflect in vicious cycles of victimisation or

perpetration of violence in adulthood.


This subsection addresses the final research question: Which institutional structures

support victims of domestic violence, and what proportion of victims are able to effectively

access available services? The analysis starts by examining the factors that may explain

why women and men who have experienced domestic violence seek (or not) external help

(subsection 4.7.1). It then looks at the public services and institutions available in Ghana to

support those affected by domestic violence (subsection 4.7.2), how many women and men

who participated in the study were aware of these services and institutions (subsection

4.7.3), and the levels of satisfaction with services among those who have used them

(subsection 4.7.4). The main aim of this analysis is to uncover information about the current

use and knowledge of and access to public services among individuals and communities,

which will be useful for future interventions aimed at preventing and mitigating domestic

violence in Ghana.

4.7.1. Seeking external help

A large international study found that most victims of domestic violence did not report

their abuse because domestic violence was often considered ‘normal’, or for fear of further

violence, losing their children or causing shame to the family (Garcia-Moreno et al., 2005).

Previous studies in Ghana had similar findings (Amoakohene, 2004). According to the

GFLHS 2015, 853 women and 419 men across the whole sample interviewed in the survey

reported having experienced at least one type of domestic violence in the 12 months

preceding the survey. Around 35 per cent of these women (297) and men (146) tried to

approach a person or organisation for help or support after experiencing violence. These

numbers were very similar to the DHS 2008 survey, which showed that 37 and 34 per cent

of women and men, respectively, who experienced violence had sought help.

Table 65 shows the share of women and men who sought external help after experiencing

any form of domestic violence. It reveals a wide variety of patterns among women and

men’s help-seeking behaviour:

Women were more likely to have sought help if they experienced physical violence

(52.8 per cent) or psychological violence (44.5 per cent) than for other forms of

violence.

212

Men were more likely to have sought help if they experienced psychological

violence (40.1 per cent) or economic violence (39.6 per cent) than for other forms of

violence.

Women were more likely than men to have sought help if they experienced sexual,

physical or psychological violence.

Men were more likely than women to have sought help if they had experienced

social or economic violence.

Table 65: Percentage of respondents who experienced domestic violence who reported

seeking external help (by sex)

Social

violence

Sexual

violence

Physical

violence

Psychological

violence

Economic

violence

Women 33.9 39.5 52.8 44.5 34.3

Men 35.6 35.1 34.3 40.1 39.6

N 531 110 242 487 508


When asked what made them seek help, the majority of women (58.5 per cent) and men

(53.3 per cent) answered that they “could not endure more”. The second most common

reason was the encouragement of friends and family (24.5 per cent for women and 28.5 per

cent for men), followed by the extent of injuries sustained (7.5 per cent for women and 8.0

per cent for men) (Table 66). These findings suggest that domestic violence is still largely

considered a private family matter: the GFLHS 2015 reported that 81.1 per cent of

respondents believed that “family problems should only be discussed with people in the

family”. Women and men who experienced domestic violence seemed to only resort to

external help in more extreme circumstances.

Table 66: Reasons for approaching external help after exposure to any form of

domestic violence by sex (in percentages)

Men Women

Encouraged by friends/family 28.5 24.5

Could not endure more 53.3 58.5

Badly injured 8.0 7.5

Afraid person/persons would kill me 0.6 1.7

The person/persons who did this to me

threatened or tried to kill me 0.9 0.5

Saw that children are suffering 0.0 0.7

Thrown out of the home 0.0 0.2

Encouraged by organisation 0.3 0.2

Other 7.4 6.3

N 146 297


213

Table 67 shows the reasons for not seeking help. The most common reason given was that

they did not think they needed help (44.1 per cent of men and 32.5 per cent of women). In

line with Garcia-Moreno et al. (2005) and the findings discussed above, 22.7 per cent of men

and 17 per cent of women did not seek help because they considered violence experienced

as ‘normal’ or ‘not serious’. In addition, 10.7 per cent of men and 16.4 per cent of women

did not seek help because they did not know who to ask – an issue that will be explored

further in the next subsections.

Table 67: Reasons for not seeking external help after exposure to any form of domestic

violence by sex (in percentages)

Men Women

Did not know who to ask 10.7 16.4

Person/organisation too far 0.2 0.1

Nobody can help me 6.9 7.6

I don’t need help 44.1 32.5

If my partner finds out it will cause trouble 0.1 1.7

If other family member finds out it will cause trouble 0.7 2.0

I don’t trust anyone 2.4 5.0

Violence normal/not serious 22.7 17.0

Embarrassed/ashamed 2.5 4.5

Afraid would not be believed 0.5 0.5

Afraid would be blamed 0.8 1.2

Afraid would end relationship 0.0 1.4

Afraid would lose children 0.0 0.3

Bring bad name to family 3.0 3.0

Other 4.6 5.7

Don't know or refused to answer 0.8 1.1

N 268 550


The qualitative data allowed the research team to explore the quantitative results further.

Data collected through focus group discussions and in-depth interviews in all ten regions

in Ghana revealed that distrust of, and poor experiences with, state authorities contributed

to individuals not seeking help. A large number of the respondents mentioned three key

challenges: the costs of the services, lengthy or ineffective delays in the handling of cases,

and corruption. Corruption, in particular, was said to affect access to justice when survivors

were given bribes and when perpetrators paid bribes or used their networks of influence to

avoid prosecution. This view was well illustrated by the following statement:

“There are times the assemblymen go to the police station to ask for the discharge of a culprit.

Normally when they have some small money on them they give it to the police in charge and

the matter will be dropped.” (FGD, married/living together men 18–50, Gbawe, Ga

South district, Greater Accra region)

214

There was a widespread view that particularly the police and the courts collected money

from those who approached them, but did not deliver appropriate care in return. This view

is exemplified by this respondent:

“For the police, they will collect money from you and would not help you solve the

problem.”(FGD, unmarried youth, Sagnarigu district, Northern region)

An opinion leader interviewed in another region added:

“People only take their case to the chief because they cannot afford the monies the police will

demand from them.” (IDI, opinion leader, Wuling, Jirrapa district, Upper West region)

Respondents also mentioned that victims were often sent back by the police to their families

to find a solution, and prosecution was discouraged:

“If a police staff knows the victim, they will ask you to take the matter to the family level for

solution.” (FGD, married/living together men 18–50, Gbawe, Ga South district,

Greater Accra region)

Concurring with this widespread perception of domestic violence as a private matter, even

within public services set up to deal with domestic violence, the majority of respondents in

the GFLHS 2015 who sought help did so in the first instance by approaching family

members or a friend (66.2 per cent of men and 77.6 per cent of women). The second source

of help was the police (sought by 14.6 per cent of men and 9.0 per cent of women as a first

point of contact) (Table 68).

Table 68: First choice of help (in percentages)

Men Women

Family member or friend 66.2 77.6

Health centre, hospital or other health

institution 1.4 1.2

Police 14.6 9.0

Queen mother 0.9 0.7

Domestic Violence and Victim Support Unit 0.6 0.7

NGO, CSO or social worker 0.6 0.2

Lawyer or court 0.6 0.7

Traditional leaders 4.9 3.0

Religious leaders 2.3 3.5

Community group 2.6 1.2

Other 4.3 2.3

Don't know or did not answer 0.9 0.0

N 145 297


The qualitative data revealed, however, that the preference of study participants for

traditional or formal legal ways of addressing domestic violence was often confounded by

215

underlying deficiencies in the justice system that, combined with structural inequalities,

may have prevented victims of domestic violence from accessing it (see also Brickell, 2015a,

2015b). As stated earlier, the 2007 Domestic Violence Act encourages mediation of all types,

but its reconciliatory approach brings some challenges. In addition to the challenges

discussed above, poorer victims and those worried about personal social stigmatisation

and reputational loss for the whole family may have preferred not to access formal justice

processes. Also, the study found that respondents were more likely to withdraw a case

from prosecution and seek customary mediation when the family or the community

determined that the victim was:

”to blame … for being the cause of the problem, when the perpetrator shows remorse [and/or]

when the family thinks about stigmatisation.” (IDI, opinion leader, Boakyekrom, Ahafo

Ano district, Brong Ahafo region)

Furthermore, respondents in focus group discussions said:

“Family and community may discourage prosecution if it is minor case like verbal assault”

or when prosecution could “lead to divorce” (FGD, married/living together women,

Garu, Garu Tempane district, Upper East region)

And, in many cases:

“…the family would convince the victim not to report the case or follow it up for …

prosecution of the perpetrator because the offender may be an opinion leader in the

community or the breadwinner of the family. Also if they like the offender, then they would

discourage his prosecution” and since “many cases of domestic violence that get prosecuted

end up in divorce, … if the couple love each other [they] would prefer to work out their

differences, the family would intervene to make sure that the union is not dissolved and so

prevent divorce and family feuds; the family and some members of the community always

try to discourage prosecution of such cases.” (IDI, DOVVSU, Gbawe, Ga South district,

Greater Accra region)

As these responses indicate, the rights of the victims are not always placed at the forefront

of customary mediation processes. Customary mediation may be strongly desired and

provide an important reconciliatory cost/time-effective approach in the Ghanaian social

context. Yet, the rights of the victim may often become submerged by the social and

economic needs of the family or community, and by prevalent gender norms. This situation

may weaken victims’ rights, while encouraging perpetrators to act with impunity.

4.7.2. Public services and institutions

The Government of Ghana has adopted a multi-institutional approach to support victims

of domestic violence after the enactment of the 2007 Domestic Violence Act. The primary

institutional response is the Domestic Violence Victim Support Unit (DOVVSU) of the

Ghana Police Service. DOVVSU acts in partnership with the Department of Social

Development, the International Federation of Women Lawyers (FIDA) and the Legal Aid

Scheme, and collaborates with the health system to provide comprehensive support to

216

victims of domestic violence. To better understand how these services currently function,

the research team conducted several observation exercises and interviews within DOVVSU

units (as detailed in Appendix C), which gathered detailed information on: (i) the facilities

available to and protocols followed by DOVVSU personnel; (ii) levels of training of

DOVVSU personnel; and (iii) the extent of the collaboration between DOVVSU and public

health services.

Facilities and protocols. The 2007 Domestic Violence Act makes provisions for police

assistance to victims and perpetrators of domestic violence. These include receiving

complaints, offering advice about victims’ rights and about services available, helping

victims to find a safe refuge when required, and providing assistance with free medical

treatment and medical evidence for possible prosecution. The interviews and observations

conducted during the study revealed that none of the DOVVSU units visited had all the

necessary facilities to offer these support services effectively. In particular, they lacked

private rooms, shelter facilities, counselling services, protection services or prosecution

advice. DOVVSU personnel interviewed in the Upper West region explained that this

situation is common across all ten regions:

“We are handicapped so if you people could help us get fresh accommodation, it will help us

very well in exercising our duties. We also need vehicles and other logistics to help us in our

line of duty.”

The research team’s visits to DOVVSU offices also revealed that, to date, the Ghana Police

Service does not provide victims’ shelters at the national or district level. As mentioned in

one of the interviews conducted with DOVVSU personnel in the Western region:

“DOVVSU have a lot of challenges […]We have issues with accommodation and cannot

house victims when the need arises. If DOVVSU had enough office space, it would have been

convenient for some victims to be brought here.”

The only CSO currently seeking to provide such a service, the Ark Foundation, receives no

support from the government.32 Most DOVVSU units consisted mainly of personnel with

limited office space. The few which had their own offices did not always have computers

to keep their records.

DOVVSU officials in all offices visited were found to be working under difficult conditions

and, sometimes, paying for the medical or transportation bills of victims themselves. The

needs of DOVVSU emerge in this poignant observation by the team at one of the DOVVSU

units included in this study:

“I saw that the police worked under difficult conditions ... they have a cage-like place near

complaints which had two compartments, men who were detained were placed in the inner

cell while women were placed in the outer one. […] I saw that DOVVSU needed a lot of

support […] as the interviewed officer told me, […] DOVVSU did not even have a car to

effect arrests. They had often had to use their little salary to foot transportation without

32 Information obtained in interviews with Mr Sandogo, Police Social Services, Director of the Ark

Foundation and National Coordinator of the Domestic Violence Coalition in Ghana.

217

getting reimbursed. The winds often blew water into their offices as it rained because the

louvre blades in the windows were either broken or not functioning anymore. This affected

the work of the officials, as they sometimes have to move from their offices taking files along

with them. Lastly, I must say this Officer was very helpful. […] The interviewed Officer was

very friendly in nature and as I was interviewing him, we were constantly interrupted by

people coming to call him out to assist them with their case at the unit.”

Other detailed observation exercises revealed that, despite being overwhelmed with their

workload and often lacking adequate facilities, the conduct of the police while dealing with

cases of domestic violence was very professional, with care being taken to file documents

appropriately (within infrastructure constraints) and to provide as much privacy and

support as possible to those seeking their help. However, the facilities available varied

considerably across regions, as follows:

Greater Accra region, Gbawe, Ga South district: There was no private room for

victims to report their cases comfortably. All the offices visited, except the

Commander’s office, looked cramped, with very little room for people to move. At

the DOVVSU desk, there were over five personnel in a very small room. There was

no fan, so the room was very hot. The room looked like an extension to the main

building; from the outside, it looked like a cage. The seats in the offices were worn

out, and many of them were old benches. Files were well stored away from the

public.

Eastern region, Kukurantumi-Koforidua: DOVVSU officials were using four

different offices. Another office, called the secretariat, was used to store all files

related to domestic violence. There was a private room where victims could sit to

file complaints.

Volta region, Gbi Kpoeta, Hohoe municipality: The police station had DOVVSU

personnel on site, but there was no separate structure for the DOVVSU unit to use.

There was only a small office that was used by the officer-in-charge. The rest of the

DOVVSU officials shared the same office with the other police officers. They

provided counselling in the same office as other police officers who could listen to

the victims’ stories. There was very limited privacy.

Northern region, Kalpohin, Sagnerigu district: The office for the entire unit

consisted of two small rooms, one of which was used by the head of the unit. The

other office was shared by about ten officers. There was limited space for people to

move about. There was no private room for victims, and reports had to be made in

front of others. However, files were kept safely out of sight.

Upper West region, Wa municipality: Office space was very small, with files stored

in a locker which looked quite old. There was no private place where victims could

report their cases. Victims had to sit in an open place where anybody could hear

them. The office did not have enough chairs and desks for the police personnel and

victims. There were no vehicles for police personnel to use in effecting arrests or to

follow up cases under investigation. The research team observed that DOVVSU

officers were not in uniform, which made them more approachable to community

members. They were very friendly with victims who approached their offices.

218

Upper East region, Garu Tempane district: The district did not have a DOVVSU unit

or staff. Interviews were conducted with district police officers in their normal line

of duty. Files were stored in a separate out-of-sight room. There was no privacy,

with people being first directed to the charge office, where an officer registered their

complaint. After listening, victims were referred to an investigator, who took their

statement. Victims were offered a seat in front of the investigator in a private room,

where he would write a detailed statement. A medical form was issued afterwards

if necessary. The victim was invited to lead the police in the arrest of the perpetrator.

If witnesses were available, victims were asked to provide them and lead the

investigator to the scene of the crime. Victims were expected to avail themselves in

court to testify.

Brong Ahafo region, Goaso, Ahafo North district: There was no building or office

which catered solely for domestic violence victims. All reports were made at the

charge office.

Central region, Cape Coast district: DOVVSU had its own separate office space, but

it was in a poor state of repair. No private room was observed.

Ashanti region, Obuasi municipality: Files were stored in a cupboard in a restricted

room at the police station. Privacy was ensured, and the procedure for filing a

complaint was the same as described above for the Upper East region.

Western region, Tarkwa district: The DOVVSU unit had two rooms. The DOVVSU

person was also the head of the police division and had to balance different duties.

The two rooms afforded limited privacy: the first room functioned also as a waiting

room and a reception area, whereas the second room was the head’s office. It was

also where files were kept.

Training of DOVVSU personnel. The establishment of DOVVSU units was intended to be

accompanied by the training of personnel to respond to domestic violence cases. The large

majority of the police personnel interviewed mentioned that they had not received

specialised training to provide help to victims of domestic violence. Neither had they

received training on gender-based violence or domestic violence, or specialised training in

listening skills to better elicit information from those seeking help. The major formal

training by the Ghana Police Service on gender-based violence took place in 2004. With the

help of UNICEF, a training manual was recently prepared to make training in gender-

based violence a basic part of the curriculum of the Police Training School; it should soon

be more widely used. DOVVSU officers were also observed to have received limited

training in information technology, and most district offices had no digital means or

training to record or access local records.

Collaboration between DOVVSU and public health services. There is close collaboration

between DOVVSU and the district hospitals which provide treatment to victims of

violence. This collaboration does not, however, include support services to victims of

violence. This is because, to enable prosecution of assault cases, medical reports have to be

completed by the medical doctor who treated the victim of domestic violence. But these

medical services are not free of charge. Many doctors request ‘unofficial’ fees from victims

219

to cover the cost of writing the medical report and the possibility of having to go to court

to explain it. This makes prosecution inaccessible to many victims, especially low-income

women. One opinion leader interviewed in the Ashanti region expressed this view, which

was common across all regions in Ghana:

“DOVVSU is really responsible for addressing domestic violence issues, but one thing

missing in their work is a centre for protection, where they will keep the victim before they

carry on with investigations. This is an issue. Secondly, when it comes to issuing a medical

report and they send it to the hospital, the doctors ask for a fee before they sign the report.

Most of these victims are vulnerable people, and they are not able to pay such fees and this

is a worry to us.”

Another interview with a medical doctor who is a domestic violence activist revealed that

many medical doctors were not trained on how to write medical reports for prosecution,

which may sometimes give lawyers of perpetrators the upper hand in reinterpreting the

reports to the benefit of their clients. Also, district hospitals do not have specialised

services, such as emergency responses to victims of violence, or priority systems for victims

of abuse unless they were brought into the hospital as normal emergency cases.

In summary, while DOVVSU is lodged within an ideal multi-institutional framework

within which to ensure effective support to victims of domestic violence, it is currently

inadequately equipped to make operational the 2007 Domestic Violence Act. At the time

of this study, DOVVSU faced the same institutional constraints that existed in the police

service prior to legal reform (see Cusack and Manuh, 2009; Ellsberg and Heise, 2005).

4.7.3. Knowledge about support services and institutions

Study respondents across all regions of Ghana showed awareness about the public services

and institutions that deal with domestic violence (Table 69). DOVVSU, however, was still

largely known by its previous name – the Women and Juvenile Unit (WAJU). The few

districts where knowledge about the public support system was more limited (in the

northern regions) were also districts where DOVVSU services were not available. In those

districts, respondents typically stated they had neither heard, seen nor used any such

public facilities to resolve domestic conflicts. But in districts where DOVVSU was

established, respondents generally included DOVVSU in their list of known public

institutions.

Table 69 provides an overview of the knowledge exhibited by the GFLHS 2015 respondents

about a range of available public services. The level of knowledge about public services

and organisations was quite high among all respondents and across all different socio-

economic groups and regions. With respect to DOVVSU, 84.9 per cent of all respondents

knew about their local units. This level of knowledge was lower among women (83.7 per

cent) than men (86.6 per cent), those living in rural areas (80.3 per cent) than those in urban

areas (89.3 per cent), those living in the Central, Volta and Northern regions (respectively,

74.2 per cent, 53.6 per cent and 67.4 per cent, compared, for instance, to 99.0 per cent in the

220

Western region) and those who had no education (71.9 per cent, compared to 98.0 per cent

for those with higher education) and were in the lowest asset quintile (70.1 per cent,

compared to 95.1 per cent for those in the top asset quintile).

221

Table 69: Respondents’ knowledge about public services (in percentages)

1 2 3 4 5 6 7 8 9 10 11 12

Sex

Male 99.9 98.9 86.6 53.8 78.9 94.9 82.6 94.7 98.4 99.6 98.7 97.9 Female 99.6 99.2 83.7 53.8 74.3 92.8 81.9 94.8 97.6 99.4 98.3 96.9

Age group (years)

15–19 99.2 99.5 80.5 52.7 71.9 92.7 78.2 95.2 99.0 99.7 98.2 97.5 20–24 99.1 98.7 85.8 56.5 78.1 94.5 82.0 94.6 96.8 98.8 98.0 97.1 25–29 99.9 98.8 87.7 53.9 78.0 93.5 84.3 94.8 97.5 99.2 98.2 97.1 30–39 99.6 99.1 86.5 55.6 76.7 93.3 83.0 94.4 97.9 99.6 98.4 96.7 40–49 100.0 99.2 81.9 50.6 73.4 94.0 81.0 94.5 97.9 99.8 98.8 98.0 50–60 100.0 99.2 83.6 52.4 76.4 93.7 82.0 95.7 98.9 99.7 99.1 98.2

Employment

Self-employed 99.7 99.0 82.3 51.9 73.9 92.5 79.2 95.0 98.6 99.5 98.2 97.2 Employed 99.8 98.8 93.0 59.6 82.6 96.1 90.3 93.6 95.6 99.1 98.7 97.2 Not working 99.6 99.4 85.7 54.6 77.1 94.7 84.0 95.1 97.9 99.8 98.9 97.6

Marital status

Never married 99.6 99.6 82.2 54.9 75.3 93.9 81.4 94.3 98.6 99.6 98.0 97.4 Married or living together 99.7 98.9 84.8 53.9 76.4 93.2 81.5 94.6 97.7 99.3 98.4 97.3 Divorced/separated/widowed 99.8 99.2 86.7 52.8 75.6 94.9 85.0 95.6 98.2 99.9 99.0 97.5

Residence

Urban 99.8 99.1 89.3 56.8 79.6 95.8 87.8 94.6 96.6 99.6 98.9 97.2 Rural 99.6 99.0 80.3 50.7 72.5 91.4 76.4 95.0 99.3 99.4 97.9 97.5

222

1 2 3 4 5 6 7 8 9 10 11 12

Region

Western 100.0 100.0 99.0 54.3 80.8 99.0 98.0 98.4 99.0 99.6 99.4 99.2 Central 99.5 98.4 74.2 25.8 54.0 96.5 84.7 96.5 98.8 99.5 99.5 97.7 Greater Accra 100.0 98.9 93.9 55.7 79.8 95.1 92.8 91.1 93.7 99.8 99.5 98.1 Volta 99.3 98.6 53.6 35.6 58.2 84.1 59.1 94.7 98.4 99.3 99.1 98.4 Eastern 100.0 100.0 96.0 72.4 92.4 99.0 98.9 100.0 100.0 100.0 100.0 99.8 Ashanti 99.9 99.5 92.6 70.1 83.3 97.1 87.9 99.1 99.5 99.8 99.5 98.6 Brong Ahafo 99.8 97.9 85.7 66.6 74.9 85.5 75.3 86.4 95.3 98.3 90.9 85.3 Northern 98.6 99.4 67.4 34.8 69.1 92.2 62.3 91.8 99.2 99.2 97.5 97.9 Upper East 100.0 97.7 84.1 21.5 72.9 85.0 51.9 91.1 99.1 99.1 99.1 99.1 Upper West 99.2 98.5 80.0 80.8 86.9 89.2 69.2 95.4 99.2 99.2 99.2 100.0

Education level

None 99.3 99.1 71.9 41.6 68.9 88.4 69.6 93.0 99.0 99.4 98.5 97.9 Primary 100.0 98.9 81.7 49.3 68.5 92.9 78.4 95.3 98.5 99.3 98.3 96.8 Middle/JSS/JHS 99.6 99.3 87.2 56.0 75.5 94.6 85.3 95.8 98.3 99.7 98.4 97.2 Secondary 100.0 99.3 90.0 61.1 84.8 96.2 86.0 94.3 97.4 99.3 98.7 97.3 Technical 100.0 97.9 91.0 56.9 77.8 95.1 89.6 95.8 96.5 99.3 99.3 97.2 Higher 99.8 98.2 98.0 65.0 91.1 97.4 94.1 94.1 94.7 99.4 98.2 97.4

Asset quintile

Lowest 99.3 98.7 70.1 43.2 65.9 88.2 67.6 92.9 99.0 99.5 98.0 97.7 Second 99.7 99.3 83.6 51.4 72.6 93.6 80.8 95.0 99.1 99.5 98.5 97.7 Middle 99.8 99.1 88.4 54.6 76.5 94.3 84.3 95.3 97.8 99.5 98.4 96.9 Fourth 99.8 98.8 90.4 59.1 82.1 95.6 89.8 96.4 96.6 99.4 98.8 97.6 Highest 100.0 99.3 95.1 63.4 86.7 97.4 91.1 94.3 96.6 99.5 98.4 96.3

Total 99.7 99.1 84.9 53.8 76.1 93.6 82.2 94.8 97.9 99.5 98.5 97.3

Source: Ghana Family Life and Health Survey (GFLHS) 2015. Notes: Service providers are 1 = Health providers, 2 = Police, 3 = DOVVSU, 4 = Shelter, 5 = NGO, CSO or social worker, 6 = Lawyer or member of Court,

7 = Legal aid group, 8 = Queen mother, 9 = Traditional leaders, 10 = Religious leaders, 11 = Religious group, 12 = Community group

223

The quantitative survey showed that despite good knowledge about public services in

general, and DOVVSU in particular, there was some room for improvement in terms of

awareness about the existence of a law against domestic violence in Ghana. According to

the GFLHS 2015, 75 per cent of respondents in urban areas and 61 per cent in rural areas

knew there was a law against domestic violence. There was also a large variation across

regions in terms of awareness about the Domestic Violence Act. For instance, only between

10.7 and 12.1 per cent of those in the Upper East, Upper West and Northern regions knew

about it.

4.7.4. Levels of satisfaction about support services and institutions

The information collected in the study has revealed that satisfaction with the public

primary response institution against domestic violence, DOVVSU, was low in comparison

with other means of support. However, very few respondents answered this question, as

not many individuals used these services. Table 70 provides an overview of satisfaction

levels with services, such as health providers, police and DOVVSU units, as well as with

traditional avenues for conflict resolution, such as traditional and religious leaders and

community groups.

Table 70: Satisfaction levels with public services (number of respondents)

Completely

satisfied

Somewhat

satisfied

Somewhat

dissatisfied

Completely

dissatisfied N

Health centre, hospital,

other health provider 10 1 0 1 12

Police 60 17 11 17 105

DOVVSU 2 1 1 2 6

Traditional leaders 19 10 2 4 35

Religious leaders 20 4 3 2 29

Community group 9 2 3 2 16


The qualitative data revealed that the reasons for the lower levels of satisfaction with

DOVVSU were mostly related to the direct and indirect costs of services, including legal

services. What bothered people was not just the financial cost but also the cost of secondary

victimisation due to lengthy procedures and additional payments:

“At the police facility, they waste a lot of time and always make demand for money so I am

really not satisfied with their service. I am however satisfied with help at health centres

because they are efficient and sometimes even treat victims of domestic violence without

referring it to the police to avoid prosecution” (IDI, opinion leader, Sagnerigu district,

Northern region)

224

Families do not always have the funds to pay for the legal costs of prosecution, which puts

them in a difficult situation:

“The main challenge has to do with the finance. Some of the clients, especially in cases of

domestic violence cannot afford the fees for their legal representation and often, we have to

take up these cases pro bono.” (IDI, opinion leader, Gbolo, Sagnerigu district, Northern

region)

GFLHS 2015 respondents were asked whether they had to pay for services when

approaching any relevant institution. The results showed that about one quarter of service

users had to pay for them. Twenty per cent of respondents who used police services paid

a fee, and about 6 per cent of respondents paid a bribe. This situation has prevented many

from accessing the services they need, and has created considerable constraints to those

who try to help and support women and men affected by domestic violence in Ghana.

225

5. Summary and discussion

This study has provided an in-depth analysis of the attitudes to and incidence,

determinants and consequences of domestic violence in Ghana, through the collection of a

rich set of quantitative and qualitative data. This section summarises the main findings of

the study in relation to each research question.

5.1. Incidence of domestic violence in Ghana

Over a quarter (27.7 per cent) of women and 20 per cent of men in Ghana

experienced at least one type of domestic violence in the 12 months prior to the

survey.

The most common form of domestic violence reported by women was economic

violence (12.8 per cent), followed by social violence (11.6 per cent), psychological

violence (9.3 per cent), physical violence (6.0 per cent) and sexual violence (2.5 per

cent).

The incidence of domestic violence was generally lower among men than among

women. The most common form of domestic violence experienced by men was

psychological violence (7.9 per cent), followed by social violence (7.7 per cent),

economic violence (7.3 per cent), physical violence (2.1 per cent) and sexual violence

(1.4 per cent).

The analysis of the qualitative and quantitative data suggested that sexual violence

was largely considered a private matter. The results above may, therefore,

underestimate the true extent of this form of domestic violence.

The GFLHS 2015 is not comparable to prior surveys of domestic violence in Ghana

because it collected data on broader and more in-depth definitions of domestic

violence than other surveys. It is, however, possible to compare the incidence of

physical domestic violence experienced by women and men in the 2008 DHS and in

the 2015 GFLHS. The results show, when using comparable categories of domestic

physical violence,33 that domestic physical violence among women decreased from

17.2 per cent in 2008 to 10.3 per cent in 2015. Domestic physical violence among men

decreased from 12.7 per cent in 2008 to 11.2 per cent in 2015.

There was a dominant view that domestic violence was largely perpetrated by men

against women. The idea of physical violence committed by women against men

was found to be ‘funny’ and sometimes ‘unimaginable’. The qualitative and

33 Please note that the estimates below for the incidence of domestic physical violence in 2015 do not match

the estimates provided in Table 26. This is because the DHS 2008 considers domestic violence to include some

categories of perpetrators that are not part of the definition of domestic relation used in this study (or in the

Domestic Violence Act 732).

226

quantitative data found, however, several instances of domestic violence committed

by women against men. For instance, 12 per cent of physical domestic violence and

37 per cent of psychological domestic violence against male household heads was

perpetrated by their spouse or (former) partner. Among the participants in the

GFLHS 2015, 789 women and 563 men reported having committed at least one form

of domestic violence, but these results must be interpreted with care due to potential

under-reporting and reporting biases.

5.2. Attitudes towards domestic violence in Ghana

Respondents overall opposed all forms of domestic violence. However, there were

important caveats as to what was perceived to constitute domestic violence.

Respondents considered that some of the definitions of domestic violence used in

the study (and in Act 732) were not violence but were, rather, accepted social norms,

defence mechanisms or ways of resolving disputes. Only physical violence between

partners was consistently perceived as violence. This is an important finding

because it suggests that violence is experienced, conceived and articulated in

different ways along a spectrum that locates some forms of violence (such as

physical violence) as the most severe, and other forms of violence (for instance,

controlling behaviour by women) as a way of managing unequal power in domestic

relationships.

Psychological violence was generally perceived as a means of defence and a coping

strategy – for women in particular – to ward off physical violence or to claim some

autonomy or agency in contexts where they may not have the ability to exert much

power beyond the use of words (insults), their body (withholding sex) or their

housework (not cooking).

Physical violence was largely considered to be unacceptable across Ghana, but some

acts of domestic physical violence were sometimes deemed acceptable, depending

on the context. There was, in particular, acceptance of wife-beating as a punishment

against disobedience and neglect of children: 11.2 per cent of men and 17.2 per cent

of women mentioned that wife-beating was acceptable if the wife had disobeyed the

husband, and 10.2 per cent of men and 16.5 per cent of women agreed that it was

acceptable if the wife neglected the children.

Attitudes of support for wife-beating have, however, become less pervasive in

Ghana since 2008: the proportion of women justifying wife-beating decreased from

36.6 per cent in 2008 to 23.1 per cent in 2015, while the proportion of men condoning

wife-beating decreased from 21.8 per cent in 2008 to 13.8 per cent in 2015.

Physical violence against children was generally viewed as unacceptable, but was

described in some discussions, particularly in the Northern and Eastern regions, as

a form of parental education to ‘steer them in the right direction’. There has been,

227

however, a significant reduction in attitudes condoning physical violence against

children: in 1998, 8 per cent of respondents thought it was never acceptable to beat

a child (Coker-Appiah and Cusack, 1999), whereas in 2015, this number rose to 25.3

per cent among men and 18.8 per cent among women.

Women were on average more likely than men to find wife-beating acceptable: 23.1

per cent of women and 13.8 per cent of men condoned wife-beating. Furthermore,

65.3 per cent of women and 56.2 per cent of men fully agreed that women were to

blame for rape if they wore revealing clothes. These findings are in line with

previous literature and reflect three facts. First, that the above forms of domestic

violence are still considered acceptable. Second, victims tend to internalise social

norms once they are in abusive relationships. Third, even when not victims, women

may internalise gender norms that hold them responsible for their experiences of

gender-based violence, rather than placing that responsibility on men or on the

broader system that reinforces gender inequalities.

5.3. Determinants of domestic violence in Ghana

The study revealed that age and intergenerational factors were strong determinants of

domestic violence in Ghana:

Young women and men (15–19 years old) were substantially more likely to have

experienced domestic violence than other age groups. Women aged 15–19 years

were between 1.3 and 4 times more likely to experience any form of domestic

violence than, for instance, women aged 30–39 years. Differences in the incidence of

domestic violence across age groups were similar for men, with the exception of

sexual violence, which was more prevalent among men aged 20–24 years. This is in

contrast with the 2008 DHS, which reported very similar incidence across age

groups, but in line with the 1998 study by Coker-Appiah and Cusack (1999).

Exposure to violence in childhood was found to be strongly related to the likelihood

of an individual being a victim of violence in adulthood. Women who witnessed

any form of domestic violence as children were between 1.3 and 3 times more likely

to report domestic violence than women who did not witness violence during their

childhood. Men who witnessed social, psychological or economic violence in their

homes as children were between 1.2 and 2.2 times more likely to have experienced

these forms of violence in adulthood than other men. This is a common finding in

domestic violence research, including the study by Tenkorang et al. (2013) on


Exposure to violence in childhood was also found to be related to the likelihood of

an individual being a perpetrator of violence in adulthood: about 28 per cent of

women and men who admitted committing social violence witnessed social

violence as children, compared to 14–15 per cent of those women and men who did

not experience violence as children. Among men who witnessed sexual violence as

children, almost 30 per cent committed sexual violence over the 12 months prior to

the survey, compared to 2.6 per cent of men who did not experience this form of

228

domestic violence during their childhood. Across psychological, physical and

economic violence, rates of perpetration are between 2 and 3 times higher for

women and men who witnessed domestic violence as children than among

individuals who did not experience domestic violence during their childhood.

Taken together, these findings suggest that Ghana may experience vicious cycles of

domestic violence reinforced by the fact that: (i) young people are more at risk of

domestic violence; (ii) experiencing domestic violence at a young age will

considerably increase the odds of experiencing domestic violence later in life; and

(iii) experiencing domestic violence at a young age will considerably increase the

odds of the individual perpetrating violence later.

Socio-economic factors were reported to be among the most important determinants of

domestic violence in Ghana, albeit in complex ways, as follows:

Low asset levels were found to be strong determinants of psychological and

economic violence among women, but high asset levels were associated with higher

levels of social and sexual violence among men.

Tensions and fights over money or property were found to be leading determinants

of all forms of domestic violence across all regions.

Women with no education were between 0.5 and 2.5 times less likely to experience

domestic violence than women with some level of education. More educated men

were around 3 times more likely to experience economic and sexual violence than

less educated men. This result may be due to the fact that women and men with

more education may also be more likely to report domestic violence, as emphasised

in previous studies conducted internationally and in Ghana.

Factors such as poverty, unemployment and economic shocks and associated

tensions were identified as important determinants of domestic violence. However,

employment status was not found to be a particularly significant determinant of

domestic violence, with some exceptions: women who were not working were 43

per cent more likely to have experienced economic violence than self-employed

women, while men who were employed or not working were, respectively, 2 times

and 2.5 times more likely to have experienced economic violence than self-

employed men.

Sexual violence was found to be more pronounced among individuals with higher

levels of education and socio-economic status, particularly men.

The study did not find a clear relationship between marital status and domestic violence.

However, the following findings are noteworthy:

Divorced, separated or widowed women were twice as likely to have experienced

physical violence than women who were never married.

Married women or divorced, separated or widowed women were, respectively, 1.8

times and 2 times more likely to have experienced economic violence than non-

married women.

Married or divorced, separated or widowed men were, respectively, 6 times and

almost 8 times more likely to have experienced sexual violence than non-married

229

men, and almost 3 times and 1.7 times more likely to have experienced social

violence than men who were never married.

Women in polygamous relationships were almost twice as likely to experience

psychological violence than women who were never married. These women were

also 3.5 times more likely to have experienced physical violence and 4.2 times more

likely to have experienced economic violence. Men in a polygamous relationship

were also more likely than men in monogamous relationships to experience sexual

or social violence.

Domestic violence sometimes followed women’s refusal to have sex with their

husbands, but this form of marital dispute was often reported to be used as a way

to retain some degree of power and autonomy in the face of other forms of violence,

especially economic violence.

There was no clear statistical relationship between geographical factors (urban/rural and

regional) and the incidence of domestic violence once the statistical analysis controlled for

other factors. This was probably because domestic violence was not necessarily related to

geographical characteristics per se but, rather, to the incidence of lower levels of

development in certain regions.

There was an association between the incidence of domestic violence and perceptions of

gender roles, patriarchal norms and masculinity. The results were particularly striking in

the qualitative data, which showed domestic violence to be closely determined by strong

gender norms that perpetuate gender inequalities. The survey also showed the following:

Women who exhibited higher levels of acceptance of wife-beating were 14 per cent

more likely to have experienced economic violence than other women. This is,

however, not a causal effect and may well suggest that victims of violence

internalise and accept gender norms that perpetuate or justify domestic violence.

Women who expressed a higher acceptance of women’s sexual autonomy were 16

per cent more likely to have experienced psychological or social violence, 17 per

cent more likely to have experienced physical violence, and 11 per cent more likely

to have experienced economic violence than women with a lower acceptance of

women’s sexual autonomy. This finding suggests that women who are more likely

to challenge accepted gender norms may be at risk of domestic violence because

they may be punished for not complying with established norms of behaviour.

Men who expressed stronger patriarchal gender norms were 26 per cent less likely

to report social violence than men who did not accept such norms.

Men who displayed higher levels of tolerance towards wife-beating were 21 per

cent, 47 per cent and 30 per cent more likely to have experienced psychological,

physical or economic violence, respectively, than men who reported lower levels of

tolerance towards wife-beating. This result may potentially reflect retaliation by

women who were beaten by their husbands.

Men who exercised more decision-making power within their households were 34

per cent less likely to experience social violence than men who exercised lower

levels of decision-making within the household.

230

The study found that exposure to criminal forms of violence in the community (theft or

robbery, vandalism, arson, kidnapping or abduction, extortion, bullying, political violence,

highway attacks, harassment, rape or defilement) was significantly associated with higher

incidence of domestic violence among men and women.

The odds of victimisation among women increased by 77 per cent for psychological

violence, 49 per cent for physical violence and 57 per cent for economic violence

when the respondent lived in a community that experienced at least one form of

violence. The effect was stronger for sexual violence and social violence, for which

the odds of having experienced this violence were 3.6 times and 1.9 times higher,

respectively, for women exposed to violence in their communities than for women

in communities that did not experience other forms of violence.

Men who reported some violence in their communities were between 2 and 3 times

more likely than men living in more peaceful communities to report having

experienced any form of domestic violence.

These findings emphasise the importance of taking into consideration factors

beyond the individual and the household when studying the determinants of

domestic violence.

Overall, the determinants of domestic violence varied considerably among types of

violence, emphasising the need for domestic violence legislation in Ghana to continue to

distinguish between different forms of violence (as in Act 732). Only young age, exposure

to domestic violence as a child and high levels of violence in the community were common

determinants of domestic violence across all types.

5.4. Consequences of domestic violence in Ghana

The GFLHS 2015 revealed adverse consequences of domestic violence on physical

health, including injuries and illness: 43.8 per cent of women who had been

physically assaulted declared having been ill in the 30 days prior to the survey – a

rate that was a third higher than among women who had not been physically

assaulted (31.2 per cent) – and 42.3 per cent of women who had experienced

psychological violence reported ill health, compared to 30.9 per cent of women who

had not experienced this form of domestic violence. The effects of domestic violence

on physical health were observed almost exclusively among women.

The results showed a very strong correlation between exposure to domestic violence

and mental health, particularly among women: women who reported serious

mental illness were 26.7, 24.2, 21.9, 19.3 and 26.3 per cent more likely to have

experienced, respectively, psychological, physical, sexual, social or economic

violence than women who did not report serious mental illness. This finding is in

line with previous studies in Ghana (Amoakohene, 2004; Adu-Gyamfi, 2014). Only

physical violence had an effect on men’s mental health status.

231

Domestic violence was found to have adverse qualitative consequences on daily life

in terms of the victim’s ability to work, go to school or do domestic work, their

ability to concentrate on daily activities, levels of confidence and feelings of living

in fear, although few respondents answered this question. The effects were larger

for women than for men, and for physical violence than for other types of violence.

Domestic violence was found to have adverse qualitative effects on child

development, educational opportunities among children and on youth deviant

behaviour, as a result of broken marriages and the resulting economic burden, and

the direct exposure of children to violence within the household – which may result

in those children being more likely to experience or perpetrate domestic violence

later in life.


Just over one in every three women and men who experienced domestic violence

approached a person or formal or informal organisation for help or support. The majority

of respondents who sought help first approached family members or a friend (66.2 per cent

of men and 77.6 per cent of women), whereas 14.6 per cent of men and 9.0 per cent of

women first contacted the police for help. These findings suggest that individuals who

experience domestic violence only resort to external help in extreme circumstances, and

that domestic violence may still be considered largely a private family matter.

Reasons for seeking help included not being able to endure the abuse further (53.3 per cent

of men and 58.3 per cent of women), encouragement by friends and family (28.5 per cent of

men and 24.5 per cent of women), and the extent of injuries sustained (8.0 per cent of men

and 7.5 per cent of women). Reasons for not seeking help included predominantly the fact

that individuals considered the violence as normal or not serious (22.7 per cent of men and

17 per cent of women), and not knowing who to ask (10.7 per cent of men and 16.4 per cent

of women).

Distrust of, and poor experiences with, state authorities and public services contributed to

not seeking help. Factors for these low levels of trust included the costs of the services,

lengthy or ineffective delays in the handling of cases, and corruption. In addition, the

Ghana Police Service had only limited logistical support to encourage the proper

functioning of DOVVSU:

In nine out of ten districts sampled, DOVVSU units or available DOVVSU staff did

not have appropriate separate offices or offer victims privacy when reporting

domestic violence. The only exception was Obuasi in the Ashanti region.

Police personnel had not received adequate specialised training to provide specific

help to the victims of violence.

Observations revealed that police personnel were motivated and professional

despite workloads and lack of resources.

232

In addition to underlying deficiencies in the justice and victims’ support systems,

preferences for traditional over formal legal ways of addressing domestic violence were

also explained by structural inequalities. Notably, poorer victims and those worried about

personal social stigmatisation and reputational loss for the whole family preferred not to

access formal justice processes and resorted to informal processes of mediation. At times,

this was imposed by the social and economic needs of the family or community, and by

prevalent gender norms.

There is room for improvement in terms of raising awareness about the existence of the

Domestic Violence Act and DOVVSU units among Ghana’s general population: 75 per cent

of respondents in urban areas and 61 per cent of respondents in rural areas knew there is

a law against domestic violence in Ghana. There was also a large knowledge variation

across regions, with 10.7 and 12.1 per cent of respondents in the Upper and Northern

regions, respectively, knowing about the law. These were also the regions where the fewest

respondents knew whether there was a DOVVSU unit at their nearest police station.

Satisfaction with the support provided by DOVVSU was low in comparison with other

means of support, but few people answered this question in the survey. Reasons for the

lower levels of satisfaction with the police and DOVVSU mostly related to the direct and

indirect costs of services, including legal services.

233

6. Way forward and policy recommendations

The key purpose of this study was to collect and analyse new evidence about the incidence

levels, determinants and consequences of and attitudes towards domestic violence in

Ghana in 2015 that can be used by the Government of Ghana, CSOs and development

partners to assess current patterns of domestic violence across the country, and to advance

legal, policy and programmatic interventions. To this purpose, this section reflects on how

existing policies and interventions, including the National Domestic Violence Policy and

Plan of Action to Implement the Domestic Violence Act, 2007 (Act 732) (2009–2019) could

be reviewed to better support victims of domestic violence and prevent domestic violence

in Ghana.

The study found that domestic violence directly affects 27.7 per cent of women and 20 per

cent of men in Ghana. In 2015, these women and men experienced at least one form of

violence within domestic relations – either social, physical, sexual, psychological or

economic violence. The causes underlying these levels of domestic violence are complex.

The study found that no single factor or process explained domestic violence victimisation

or perpetration – these varied considerably across the different types of domestic violence

and depended on (i) personal and socio-economic characteristics of victims or perpetrators;

(ii) specific relationship dynamics within families; (iii) norms and structures that operate at

the level of the household, the community and the wider society; and (iv) national- and

global-level factors that shape prevailing gender and social norms, and access to resources

and opportunities by different individuals and social groups.

To address these complex, intersecting factors, policy responses must operate at several

levels and in several sectors. This section discusses which potential interventions may work

– and how existing interventions and programmes may be strengthened – to reduce the

incidence and impact of domestic violence in Ghana. The section is organised around the

main sections of the analysis, and discusses potential recommendations with respect to the

incidence of domestic violence (subsection 6.1), attitudes towards domestic violence

(subsection 6.2), determinants of domestic violence (subsection 6.3), consequences of

domestic violence (subsection 6.4) and use of and access to support services and institutions

(subsection 6.5).

As with other studies, this study uncovered intriguing findings that could not easily be

explained within the scope and time frame of the project, and generated several additional

questions. This section concludes, therefore, with recommendations for future research

work on domestic violence in Ghana. These recommendations focus on four important

areas that were outside the scope of this study: (i) the incidence, causes and consequences

of non-domestic forms of violence, which were found to affect considerable numbers of

women and men in Ghana; (ii) the analysis of causal effects with respect to the causes and

determinants of domestic violence; (iii) attitudes towards and the incidence, causes and

consequences of sexual violence, particularly among men; and (iv) attitudes towards and

234

the incidence, causes and consequences of domestic violence among children, which was

uncovered as an area of concern for future research and policy work.

6.1. Recommendations with respect to the incidence of domestic violence in

Ghana

1. Amendment of the Domestic Violence Act, 2007 (Act 732) to include social abuse as a

specific dimension of domestic violence in the law. The study has shown that social

violence is the second most predominant form of domestic violence against men and

women. It is therefore recommended that the Domestic Violence Act, 2007 (Act 732) is

amended to specifically include this category of abuse within the categories of prohibited

forms of domestic violence in Section 1, in addition to covering physical abuse, sexual

abuse, economic abuse and emotional abuse, as well as harassment. Currently under

Section 1(c) of the Domestic Violence Act it is indicated that domestic violence includes

“harassment including sexual harassment and intimidation by inducing fear in another

person”. In Section 42 of the Domestic Violence Act ‘harassment’ is defined as including

acts of intimidation, threatening acts and some aspects of controlling behaviour. However,

not all the forms of social violence found in the course of the study are set out or obvious

in this definition. It is, therefore, important to specifically include social violence in the

definition of violence in Section 1. The working definition of social violence for the study

was “acts of controlling behaviour experienced by men and women, such as preventing

someone from seeing friends or family of birth; stopping someone from leaving the house;

insisting to know where someone is at all times; stalking; spreading false information,

videos or photos without permission; or forcing women to have an abortion”. This

definition could be adopted in any further amendments of the law.

2. Extend gender-sensitive legislation and awareness and advocacy programmes to bring

attention to non-physical forms of domestic violence. The study has shown that people’s

lived experiences of domestic violence are sometimes at odds with the way in which

domestic violence is conceptualised by policymakers and researchers. This may undermine

effective, tailored policy responses at a practical level. For example, many acts defined as

psychological or economic violence in the Domestic Violence Act were not perceived as

violence by the respondents. These perceptions may result in a mismatch between

legislation and practice, whereby victims of these forms of violence do not feel they ought

to seek support, leading to fewer meaningful policy responses to less overt and less

physical forms of violence. Addressing these issues will involve taking a more

comprehensive approach to gender-responsive legislation in Ghana through gender-

sensitive legislative frameworks that may enable state institutions to develop responsive

policies and transformative protocols to challenge all forms of (physical and non-physical)

domestic violence. These measures should be combined with awareness-raising

campaigns targeting the public, media, police, health, legislators and other decision-

makers and stakeholders about the importance of non-physical forms of domestic violence.

235

3. Improve support available to victims of psychological domestic violence. The study

found that psychological violence was the most common type of violence experienced by

men (affecting 7.9 per cent of men) and the third most common form of domestic violence

experienced by women in Ghana (affecting 9.3 per cent of women). Psychological violence

was also found to be associated with a high risk of women suffering from serious mental

illness. At the same time, psychological violence was generally perceived as a means of

defence and a coping strategy – for women in particular – to ward off physical violence or

to claim some autonomy. These findings required a multi-component approach including

awareness-raising campaigns about the importance and prevalence of psychological

violence, further recognition in support services about this form of violence, strengthening

the skills of mental health professionals to identify and offer protection to those at risk of

this form of abuse (such as psychotherapy and counselling services), and further

cooperation between mental health professionals, NGOs and police units involved in

domestic violence prevention.

4. Further attention is needed to address the widespread prevalence of economic violence.

Economic violence was defined as the denial of household expenses money (chop money)

even if enough financial means are available; unsolicited taking of money; control of

belongings and spending decisions; damage to or destruction of someone’s property;

denial of the right to work; forcing someone to work against their will; or denial of food

and other basic needs. Economic violence was the most prevalent form of domestic

violence against women (affecting 12.8 per cent of women) and the third most common

form of domestic violence against men (affecting 7.3 per cent of men). Economic violence

was particularly prevalent among married women, and less wealthy men and women

living in rural areas, and was found to have severe consequences for the physical and

mental health status of women. These findings call for a stronger coordination of the laws

and policies relating to abuse of individual property rights within domestic relations, and

to economic maintenance provision for married, separated or divorced individuals. There

are provisions on property rights of spouses in Article 22 of the 1992 Constitution and the

Matrimonial Causes Act, 1971 (Act 367), and, apart from the Domestic Violence Act, there

are provisions on economic violence in the Criminal Offences Act, 1960 (Act 29). There will

also be a need to better coordinate domestic violence support services with ongoing

interventions aimed at improving the economic conditions of women (such as

microfinance and cash transfer programmes). Better coordination may ensure a reduction

of the economic dependence of women on men – a factor found to be central to explaining

the prevalence and consequences of domestic economic violence.

5. Extend education, awareness and advocacy programmes to bring attention to domestic

violence against men. The study showed that, although women are more at risk of

experiencing domestic violence, a substantial number of men are also at risk of violence

within domestic relations. Addressing this issue will require bringing attention to forms of

domestic violence experienced by men, and targeting awareness-raising campaigns

towards the value of understanding how men also struggle with harmful masculinities

that might be imposed on them by other men and women. There is also a need to shift

236

social norms about the acceptance of different forms of violence among women – and men.

Further training of police and other support services may also be required, as men find it

even more difficult than women to report cases of domestic violence (when the response

may be disbelief, or ‘giggling’ as some of the qualitative transcripts show).

6. Further attention is needed with respect to child protection. Some forms of physical

violence against children are sometimes perceived as a form of parenting. Non-biological

children can be at particular risk of domestic violence, particularly when the mother is in

a vulnerable economic position. This calls for a policy decision to ensure protection for

children, especially vulnerable non-biological children. It may also call for health and

education workers who engage with children most frequently to receive specialised

awareness training on indicators of domestic violence and the ways available to ensure

that children are protected. Interventions aimed at schools, teachers, children and parents

have reduced violence against children in other countries (see Fulu et al., 2014). Examples

of such interventions include work in schools and with teachers to raise their awareness

about violence-affected children, and improve their skills to behave in non-violent ways;

work with schools and children through curriculum-based awareness-raising

interventions; and work with parents to build skills that involve non-violent means of

parenting. This is, however, an area not explored in detail in the study (the study did not

involve children as participants) and requires further research.

6.2. Recommendations with respect to attitudes towards domestic violence in

Ghana

1. Need for new sensitisation programmes about harmful social norms that drive domestic

violence. The study found that attitudes against any form of domestic violence were strong

and widespread, but many of the definitions of domestic violence used in the study (and

in Act 732) were seen by respondents not as violence but, rather, as accepted social norms,

defence mechanisms or ways of resolving disputes. Only physical violence between

partners was consistently perceived as violence. Psychological violence was generally

perceived as a means of defence and a coping strategy – for women in particular. Other

acts defined as domestic violence in the 2007 Domestic Violence Act – such as controlling

behaviour, verbal abuse or men making sexual remarks to women – were sometimes

deemed acceptable. In particular, wife-beating was accepted as a valid punishment against

disobedience and neglect of children, and rape was considered by the majority of

respondents to be the fault of the woman if she wore revealing clothes. There was also

limited discussion about sexual violence in all focus group discussions and individual

interviews, reflecting the fact that many acts defined as domestic sexual violence may

either be viewed as private or not perceived as violence but, rather, as ‘normal’ behaviour.

These findings suggest that advocacy and awareness-raising campaigns (perhaps through

the use of the media) still have a strong role to play in working to change these widely

accepted norms and behaviours that justify the violence associated with them.

237

2. Need for new sensitisation programmes to reduce acceptance of domestic violence,

particularly among women and girls. The study found that women were on average more

likely than men to find some forms of domestic violence more acceptable. These included

wife-beating, appropriateness of sexual remarks against women, and rape under some

circumstances (if women wore revealing clothes). These findings reflect two important

facts that shape the prevalence of domestic violence in Ghana. First, women who are

victims of domestic violence may have internalised the acceptability of domestic violence

once they are in abusive domestic relations. Second, even when not victims, women may

internalise gender norms that hold women responsible for their experiences of gender-

based violence, rather than placing that responsibility on men or on the broader system

that reinforces gender inequalities. The prevalence of harmful gender norms – such as

discouraging women from engaging in paid labour, expectations that men are the

breadwinner, forcing women to obey men or blaming and shaming women for violence

inflicted on them – was found throughout the study as central to explaining the prevalence,

causes and consequences of domestic violence in Ghana. For instance, the regression

analyses conducted in the study showed that women who expressed a higher acceptance

of women’s sexual autonomy were at greater risk of experiencing psychological, social,

physical or economic violence. The research team interpreted this finding as indicating that

women who challenge established gender norms may be at risk of domestic violence

because they may be punished for their non-compliance with established norms of

behaviour. These findings call for multi-component interventions and strong cooperation

between state agencies and civil society to change the prevalence of harmful gender norms

that perpetuate gender inequalities and condone several forms of domestic violence.

Examples of interventions include (i) economic interventions to reduce the economic

dependence of women on men (such as microfinance, cash-for-work, or cash transfer

programmes); (ii) interventions that work with men and boys to reduce the prevalence of

harmful masculinities and harness them to bring about positive change; (iii) interventions

that work with girls and women to improve their confidence and sense of worth (perhaps

also including psychotherapy); (iv) interventions at the level of family relationships (such

as counselling); and (v) group education campaigns and mobilisation activities at the level

of communities (involving men and women, boys and girls) to raise awareness about the

prevalence of these gender norms and find collective ways to reduce or abandon such

norms altogether.

3. Further awareness-raising is needed about physical violence against men and children.

The qualitative data found a close relationship between participants’ knowledge of legal

rights and acts, such as the Domestic Violence Act, and their perception that physical forms

of domestic violence, especially against women, were unacceptable. Physical violence was

generally interpreted as committed by men against women. Discussions about perceptions

of physical violence committed by women against men revealed that for many people this

was ‘funny’ and sometimes ‘unimaginable’. Physical violence against children, while often

perceived as unacceptable by both men and women, was described in some focus group

discussions, particularly in the Northern and Eastern regions, as a form of education. This

calls for more education and awareness-raising campaigns about the fact that physical

238

violence against anyone, including children, is unacceptable. Interventions discussed

above with respect to the prevalence of sexual violence among men and violence against

children would be useful to increase awareness about the prevalence of attitudes that

condone physical violence against men and children. The Northern and Eastern regions in

particular should be targeted for further interventions in this latter area.

4. Need to target further sensitisation programmes against domestic violence in less

economically developed areas. The study has shown that attitudes against domestic

violence were consistently less pronounced in rural and in the less economically developed

northern areas, with individuals in the lowest asset quintiles and with no education

consistently being more likely to condone several forms of domestic violence than other

socio-economic groups. This finding indicates that sensitisation programmes should target

messages in less developed areas and among populations with lower economic status to

enable a change in attitudes and norms that reinforce domestic violence. The findings also

call for increased awareness-raising in rural communities to ensure an elimination of the

acceptance of abusive behaviour in inter-personal relations and within the domestic

setting. This is particularly important given the intergenerational dynamics of domestic

violence found in this study, and the value of stopping the vicious cycle that adversely

affects people who experience domestic violence at a young age.

6.3. Recommendations with respect to the determinants of domestic violence in

Ghana

1. Need to target interventions towards adolescents and young adults. The findings

indicate strong effects of age and intergenerational factors on the incidence of domestic

violence in Ghana. Young women and men (15–19 years old) were substantially more at

risk of domestic violence than other age groups. Exposure to violence in childhood is also

strongly related to the likelihood of being a victim and perpetrator of violence in

adulthood. These findings suggest that Ghana may experience vicious cycles of domestic

violence reinforced by the fact that young people are more exposed to domestic violence

and that experiencing domestic violence at young ages considerably increases the odds of

further exposure to and perpetration of violence. This calls for targeted interventions,

including sensitisation and therapy for individuals who were exposed to violence at a

young age, to stop the cycle of violence, and coordination with the interventions included

above in Section 6.1 with respect to the incidence of violence against children.

2. Further attention is needed on polygamous relationships. Even though the study found

no clear relationship between marital status and domestic violence, it found that women

in a polygamous relationship were 1.9, 3.5 and 4.3 times more likely to experience

psychological, physical or economic violence, respectively, than never-married women.

Men in polygamous relations were also the most likely to experience sexual or social

domestic violence. This finding emphasises the need to strengthen legislative systems

against polygamy, which violates Article 17 of Ghana’s 1992 Constitution, or to ensure that

239

individuals in polygamous relations are protected under legislation against domestic

violence.

3. Reinforcement of laws and practices about economic maintenance. The study found that

increases in domestic violence were associated with (women’s) refusal to have sex. This

refusal was often used as retaliation against other forms of violence, especially economic

violence. In addition, shirking of economic responsibilities, especially by men not

providing chop money, not paying children’s school fees or not paying other bills, was seen

as an important determinant of domestic violence. In addition to provisions against

economic violence in the Domestic Violence Act, the Children’s Act, 1998 (Act 560) has

provisions that can be used to hold parents responsible to provide for their children. It is,

therefore, important to enforce these laws alongside each other. A recommendation is also

made to support the proposal by the MoGCSP to set up a Child Support Unit to hold non-

custodial parents responsible for contributing to the maintenance of their children, as set

out in the Justice for Children Policy. On the issue of the insistence on the provision of

daily chop money by men for the upkeep of homes, there is a need to work on changing

gender norms, as discussed above, that place women as economically dependent on men,

and expect men to be the sole provider for the family. There may also be the need to explore

whether the expectation for men to provide chop money is simply a means by which men

are made to contribute to housekeeping costs. Encouraging community dialogue (between

men and women) that promotes women’s economic participation, and that values

women’s unpaid labour in the household, might also promote a greater awareness of the

important roles that women can play in the household, alongside men.

4. Reinforcement of property laws. Fights over money, assets or property were found to be

prominent determinants of domestic violence across all regions. Economic tensions were

mentioned to lead to multiple forms of domestic violence among adults – including

between partners, siblings and in-laws. There should be a concerted effort to ensure that

laws in place that protect the property rights of spouses on the death of a spouse – such as

the Intestate Succession Act, 1985 (PNDCL 111) – are disseminated and enforced. With

regard to the property rights of spouses, the Property Rights of Spouses Bill, 2013 that seeks

to actualise Article 22 of Ghana’s 1992 Constitution, which is before Parliament, should be

passed into law as a matter of urgency to protect the property rights of married people at

the dissolution of their marriage or the separation of cohabiting people. The police should

be made aware of these laws to ensure that they are enforced alongside the Domestic

Violence Act.

5. Need to continue sensitisation programmes about alcoholism. Alcoholism was

frequently mentioned as one particularly problematic form of harmful behaviour

associated with domestic violence. Some domestic conflicts were shown to be triggered

following excessive alcohol intake. There is a need for policymakers to continue to educate

people about the harms caused by alcoholism and provide support for the treatment of

alcoholism. There is also a need to advocate that the use of alcohol is not an acceptable

240

excuse or justification for acts of domestic violence and will not stop prosecution for

violations committed.

6. Reinforce legislation and sensitisation interventions against early child marriage. Early

child marriages can be considered a form of domestic violence perpetrated by families

against girls with the consent of both parents. The qualitative data showed evidence of this

practice in some northern regions. Even though the Children’s Act prohibits forced and

early marriage, it is a practice that goes on in some parts of Ghana. A recommendation is

made that the MoGCSP continues undertaking research into the issue and formulates

appropriate policies to address it, including prohibiting it as well as including it in the

Domestic Violence Act.

7. Need to look more closely at the association between domestic violence and community

violence. The results of the study showed a strong association between criminal violence

in the community and incidence of domestic violence. Men who reported violence in their

communities were two to three times more likely than others to report any form of

domestic violence. The odds of women experiencing any form of domestic violence in

communities that experienced other forms of criminal violence increased by 77 per cent for

psychological violence, 49 per cent for physical violence and 57 per cent for economic

violence. The effect was even stronger for sexual violence and social violence, for which

the odds of domestic violence were 3.6 times and 1.9 times higher, respectively, for women

living in communities that experienced other forms of violence, in comparison to women

living in other communities. The qualitative data emphasised in addition that in the

northern regions of Ghana chieftaincy disputes can erupt into community-wide violence.

This finding calls for a national policy to address community violence generally, and

violence associated with chieftaincy and other traditional forms of leadership in specific

places, since this may result in increased incidence of domestic violence. This requires

coordination between different parts of the Ghanaian police dealing with different forms

of violence across communities.

6.4. Recommendations with respect to the consequences of domestic violence in

Ghana

1. Strengthen the involvement of medical practitioners in identifying victims of domestic

violence. The study showed that domestic violence had several adverse consequences for

the physical health of women. These included injuries (particularly because of physical

violence), increased blood pressure and general illnesses. This finding suggests the need

for further cooperation and integration of police and medical practitioners in the

identification and reporting of victims of domestic violence, since the Domestic Violence

Act allows medical practitioners to lodge complaints of domestic violence on behalf of

victims. There is also the need for sensitisation programmes to be undertaken with medical

practitioners.

241

2. Strengthen mental health policy in relation to domestic violence. The results showed a

very strong correlation between exposure to domestic violence and mental health,

particularly for women, and between domestic violence and the inability to perform daily

tasks, feeling depressed and being unable to cope on a daily basis. This finding suggests

that stronger linkages should be forged between domestic violence and mental health

policies. It also calls for stronger and more effective inter-agency collaboration to address

the incidence and harmful consequences of domestic violence. This may include targeted

interventions such as sensitisation and therapy for individuals who experienced domestic

violence, including those who were exposed to violence at a young age. It should also

include supporting the use of psychologists under the Domestic Violence Act and making

the cost of accessing their services affordable to victims and perpetrators.

3. Strengthen child protection policy in relation to domestic violence. The study found that

domestic violence adversely affected child development, educational opportunities among

children and youth behaviour. Domestic violence was also found to have negative impacts

on the education and economic potential of children. This finding suggests that the

MoGCSP must intensify its work around child protection. It suggests also the need for

stronger and more effective inter-agency linkages and strategies to address child protection

more effectively across its various dimensions. There is a particular need to incorporate

information about the adverse consequences of domestic violence in education, awareness-

raising and advocacy programmes, and in work with schools and parents, as effects of

domestic violence go beyond the victim (for instance, on children and on family stability)

and are not widely acknowledged.

6.5. Recommendations with respect to the use of and access to victims’ support

services and institutions in Ghana

1. Reinforce sensitisation programmes about the need to report all forms of domestic

violence. The findings indicated that only about one third of women and men who

experienced at least one form of domestic violence in Ghana in 2015 tried to approach a

person or organisation for help or support. Reasons for seeking help include not being able

to endure the abuse further, encouragement by friends and family, and the extent of

injuries sustained. Individuals exposed to domestic violence seemed to only resort to

external help under more extreme circumstances, and the majority of respondents who

sought help approached family members or a friend first. Reasons for not seeking help

included considering the violence ‘normal’, not knowing who to ask, lack of trust, and

feelings of embarrassment and shame. These findings suggest that domestic violence is still

considered largely a private matter. There is a need for more education and sensitisation

to make people aware that domestic violence is an offence, and not a private matter, and

that this applies to all forms of violence under Act 732. People should be encouraged to

report the occurrence of all forms of domestic violence and should receive adequate

support when turning to formal institutions.

242

2. Improve awareness of the 2007 Domestic Violence Act and available services,

particularly in rural and northern areas. Awareness about domestic violence has improved

since previous studies in Ghana. However, 25 per cent of respondents in urban areas and

39 per cent of respondents in rural areas were not aware of the law against domestic

violence in Ghana. Television and radio were found to be good means of transmission of

information in urban areas (48.3 per cent of urban people were made aware of the Act via

television, and 67 per cent via radio). These means were less effective in rural areas (19.3

per cent of rural respondents knew about the Act via television, and 54.6 per cent via radio).

Building further collaboration between DOVVSU personnel and specialised CSOs, public

units, including the Information Services Department and the National Commission for

Civic Education, media and resource persons in local communities may strengthen

institutional responses to domestic violence. There also seems to be a large variation across

regions in terms of awareness about the law. For instance, 66 per cent of respondents overall

knew that there is a law, in contrast to 10.7–12.1 per cent of respondents in the Upper East,

Upper West and Northern regions. In these regions, very few respondents were also aware

of the presence of a DOVVSU unit. Further efforts are needed to improve awareness of

existing legislation and support channels in these areas.

3. Increase the number, spread, awareness and effectiveness of DOVVSU units. The study

has shown that there is a need for more service units. Only a quarter of survey respondents

in rural areas and 36 per cent in urban areas mentioned that the nearest police station had

a DOVVSU unit. More DOVVSU units across the country would also reduce travel burdens

for victims of domestic violence. It is, however, important to note that part of the reason for

the low usage and knowledge of DOVSSU had also to do with people perceiving traditional

mediation as more acceptable, cheaper and less likely to result in social shaming. The

NPPOA has made provisions for coordination by the DV Secretariat to ensure that various

actors, including CSOs, traditional authorities and others involved in addressing domestic

violence issues, play a role at the community level and report to the regional and national

levels, and there is the need for the Domestic Violence Management Board to ensure that

this coordination system is actualised. It is important that formal state institutions

collaborate with more informal community actors in settling disputes, documenting

settlements and referring matters where necessary – and also in ensuring that different

actors do not further reinforce existing social norms and behaviours that restrict action

against and reporting of domestic violence. There is, therefore, an urgent need to ensure

that the implementing regulations (legislative instruments) that will roll out the Domestic

Violence Act are enacted as soon as possible to implement the Act further, including how

to apply for various protection orders, qualifications required for providing psychological

services, minimum standards for setting up shelters and developing training manuals, and

how to access the Domestic Violence Victim Support Fund.

4. Increase the capacity of service providers to ensure privacy in units and provide

additional support services, such as counselling and shelters. People who experience

domestic violence may have immediate needs beyond medical care, such as access to

therapeutic services, accommodation, legal aid and economic opportunities. Psychological

243

assistance is also often needed by victims of domestic violence who struggle to address

feelings of shame and guilt. The Domestic Violence Act makes provision for these, and the

regulations of the Act are further expected to elaborate on how to access these services. It

is important that the regulations make clear the need for helpline support and information,

shelter/refuge and safe accommodation, short- and long-term psychological counselling,

legal advice, advocacy and outreach services.

5. Build the response capacity of specialised personnel. There is a need to update and

improve the training of police officers to support women and men, boys and girls, as

recommended in the NPPOA. The study found that the staff involved in domestic violence

support services were highly motivated but lacking in wider institutional support.

Additional training and support should enable them to better support the needs of victims

of domestic violence. Consistent training should be encouraged and made part of the short-

and long-term budget of the MoGCSP. Training should also include awareness about

current research findings on the prevalence and gendered dimension of violence globally,

regionally and locally in Ghana. This will encourage the design of appropriate tools to help

specialised support units detect domestic violence at the first contact with women and men

who ask for help, as either victims or perpetrators. It would also help support units to

approach domestic violence respectfully and sensitively.

6. Improve trust in public services that support victims of domestic violence. Distrust of

health care, police and legal services poses a challenge to support services across all regions

of the country. While state institutions may be providing resources to effectively implement

the Domestic Violence Act, they need to engage more directly with how resources are

allocated at a local level and improve the public’s trust in the way in which domestic

violence cases are handled and processed. At a practical level, a free telephone helpline for

people to call in and report cases might be useful. The government could also consider

setting up incentive programmes to encourage police and health care workers to report

cases of domestic violence and follow appropriate procedures. The enactment of the

regulations of the Domestic Violence Act will also strengthen the work of public services

that support victims of domestic violence.

7. Improve inter-agency collaboration. The study observed that district hospitals do not

have specialised services, such as emergency response to individuals who experience

domestic violence, or priority responses for victims of abuse – unless victims are brought

in as normal emergency cases. The MoGCSP should facilitate the development of inter-

agency collaboration between service providers to address this gap so that there is better

collaboration between medical facilities, the police, the judiciary and other institutions

mentioned in the NPPOA. The MoGCSP should also collaborate with the Ministry of Health

to ensure that special spaces are created in district and regional hospitals for victims of

domestic violence.

244

6.6. Recommendations for future research on domestic violence in Ghana

1. Explore further the incidence, determinants and consequences of non-domestic forms of

domestic violence in Ghana. The main aim of this study was to document and analyse the

underlying attitudes towards and incidence, norms, determinants and consequences of

domestic violence in Ghana – i.e. violence that takes place within domestic relations as

defined by the 2007 Domestic Violence Act. The GFLHS 2015 collected, however, a

substantial amount of information on social, physical, sexual, psychological and economic

violence that takes place in non-domestic settings, such as the workplace and the

community. The study found that a large percentage of women and men experienced non-

domestic forms of violence. This is an interesting finding that deserves further exploration

for two main reasons. First, the large incidence of forms of gender-based violence outside

domestic relations suggests that harmful gender norms that condone the use of domestic

violence may also be highly prevalent in the wider society. Therefore, policies that address

violence within domestic relations may miss out vulnerable men and women who continue

to experience forms of violence outside domestic relations. Second, the persistence of

gender-based violence outside domestic settings suggests the widespread acceptance of

gender norms that may contribute to the entrenchment and acceptance of forms of

domestic violence as ‘normal’. The findings of this study suggest an urgent need to explore

further the incidence, determinants and consequences of gender-based violence outside

domestic settings in Ghana – and the data collected in this study could provide the baseline

for such future research.

2. Further research on establishing causality in the determinants and consequences of

domestic violence in Ghana. The multivariate regressions discussed in Section 4 of this

report allowed this study to provide important information about the determinants and

consequences of domestic violence in Ghana that took into consideration potential

correlations between individual variables. This is important because policy action based

on simple cross-tabulations may be misleading, as simple correlations may miss important

interactions between the different complex factors associated with the incidence and

consequences of domestic violence among some groups of people or some regions.

However, the multivariate regressions analysed in this study are only a small step towards

identifying precisely the determinants and consequences of domestic violence. The

multivariate regressions produced in the study provided statistical relationships that take

into account complex interactions between variables but do not allow the identification of

precise causal relations. Establishing the direction of causality among the various factors

that shape domestic violence and among the consequences of domestic violence requires

the use of sophisticated econometric methods that were outside the scope of this study due

to budget and time restrictions.

One example regards findings about the role of education as a determinant of domestic

violence. The study found that the risks of social or sexual violence were higher for men

with higher education levels. This interesting result could be due to a variety of causes. For

instance, it is possible that the risk of domestic violence increases for men with higher

245

education levels because they may have more confidence to report domestic violence. In

this case, domestic violence is higher among this group not due to education per se but

due to other social factors that jointly explain education levels and reporting domestic

violence (for instance, more self-confidence). The analysis provided in this report is not

sufficient to establish the causal reasons that explain the overall finding.

Another example is the relationship reported in the study between violence at the

community level and incidence of domestic violence. The analysis conducted in the study

revealed a strong and statistically significant relationship between these forms of violence

but was not able to identify the direction of causality. It is possible that communities that

experience high levels of community violence may also experience high levels of domestic

violence because violence becomes an accepted norm of behaviour. However, it is also

possible that experiencing domestic violence may predispose individuals to other forms of

violence. These two examples illustrate two statistical challenges – called, respectively,

omitted variable biases and reverse causality – that can only be solved with more

sophisticated quasi-experimental techniques. Efforts should be made to advance causal

research on the determinants and consequences of domestic violence, as this would

uncover more precise entry points for interventions aimed at preventing and remedying

domestic violence and their rigorous evaluation. The datasets collected in this study can

be used for this purpose in future research.

3. Further research on the incidence, causes and consequences of and attitudes towards

sexual violence, particularly against men. This study found that domestic sexual violence

exhibited patterns that were different from all other forms of domestic violence. Notably,

it was more prominent among employed, better-off, less economically dependent

individuals living in urban areas, particularly men. Domestic sexual violence against men

– as well as overall sexual violence over their lifetime and in the 12 months prior to the

GFLHS 2015 – was more common among men in the top asset quintile and with higher

education than among men with lower levels of assets and education. The association for

women was less clear, but there was evidence that better educated women are more likely

to have experienced overall sexual violence (within and outside domestic relations) over

their lifetime and in the 12 months prior to the survey than women with lower levels of

education. Furthermore, while most acts of sexual violence against women were

predominantly perpetrated within domestic relations, most acts of sexual violence against

men were predominantly perpetrated outside domestic relations. It was difficult to

provide a clear explanation for these results without the use of more sophisticated

econometric techniques.

The research team postulated that one likely explanation for these results may be that these

individuals are more likely to have reported these forms of violence, which remain hidden

across most of Ghana. The results discussed in subsection 4.2.3 on attitudes towards

domestic sexual violence provided some evidence that may support this explanation. First,

the study found that there was very limited discussion about sexual violence in the

qualitative data. Second, most survey respondents showed high levels of acceptance of

246

behaviours that the study defined as sexual violence. Taken together, these findings

suggest that sexual violence may be understood largely as something that takes place

outside domestic relations. Perceptions of what may constitute sexual violence within

domestic relations seemed very limited – or hidden. However, much more research is

needed to better understand: (i) the true incidence of domestic sexual violence; and (ii) the

reasons why sexual violence, both domestic and outside domestic relations, was more

prominent against wealthier, more educated men. The research team was also not able to

relate the incidence of this violence to the place where it may take place – for instance, their

workplace, places where they socialise or within the community. There is a need for future

research that makes use of the GFLHS 2015 data to analyse in more depth, using more

sophisticated statistical techniques, the reasons underlying these findings, and also future

studies that analyse in more detail the incidence, causes and consequences of domestic

violence – but also the places where it takes place and the behaviours that may shape its

incidence among particular population groups.

4. Further research on the exposure of children to domestic violence. This study focused on

the analysis of the incidence, determinants and consequences of domestic violence in

Ghana among people aged between 15 and 60 years old. However, the findings reported

in the study – particularly in the qualitative data – highlight the exposure of younger

children to domestic violence. This exposure takes the form of direct victimisation, when

children are beaten or abused in other ways as part of parental disciplining, or indirect

exposure to violence committed against others within their homes. The quantitative data

has shown, in addition, that the adults surveyed in the GFLHS 2015 are more likely to

experience domestic violence if they were exposed directly or indirectly to domestic

violence when they were children. These findings suggest that some families and

communities in Ghana may experience the transmission of norms and attitudes across time

and generations that reinforce the persistence of domestic violence. Addressing this issue

will require more knowledge about the incidence, determinants and consequences of

domestic violence among children under the age of 15. This research was outside the scope

of this study. However, the findings indicate an urgent need for further research in this

area.

247

References

Abane, H. (2003). “For Better, For Worse: Social Dimensions of Marital Conflict in

Ghana:The Case of Cape Coast”, Gender and Behaviour 1: 34–54.

Abramsky, T., Watts, C.H., Garcia-Moreno, C., Devries, K., Kiss, L., Ellsberg, M.,

Jansen,H.A. and Heise, L. (2011). “What Factors Are Associated with Recent

Intimate Partner Violence? Findings from the WHO Multi-country Study on

Women's Health and Domestic Violence”, BMC Public Health 11: 109.

Addai, I. and Adjei, J. (2014). “Predictors of Self-Appraised Health Status in Sub-Saharan

Africa: The Case of Ghana”, Applied Research Quality Life 9: 233–53.

Adinkrah, M. (2014). “Homicide–Suicide in Ghana Perpetrators, Victims, and Incidence

Characteristics”, International Journal of Offender Therapy and Comparative Criminology

58 (3): 364–87.

Adjetey, F.N.-A. (1994). “Reclaiming the African Woman’s Individuality: The Struggle

Between Women’s Reproductive Autonomy and African Society and Culture”, Am.

UL Rev., 44: 1351.

Adomako Ampofo, A.A. (2008). “Collective Activism: The Domestic Violence Bill

becoming Law in Ghana”, African and Asian Studies 7: 395-421.

Adomako Ampofo, A.A. (1993). “Controlling and Punishing Women: Violence against

Ghanaian Women”, Review of African Political Economy 20 (56): 102–11.

Adu-Gyamfi, E. (2014). “Challenges Undermining Domestic Violence Victims’ Access to

Justice in Mampong Municipality of Ghana”, Journal of Law, Policy and Globalization

27 (0): 75–90.

Africa, A. (2010). “Standpoint:“Murderous Women”? Rethinking Gender and Theories of

Violence”, Feminist Africa 14 Rethinking Gender and Violence.

Aizer, A. (2011). “Poverty, Violence, and Health: The Impact of Domestic Violence during

Pregnancy on Newborn Health”, Journal of Human Resources 46 (3): 518–538.

Aizer, A. (2010). “The Gender Wage Gap and Domestic Violence”, American Economic

Review 1847–1859.

Akpalu, E.A., Ofei-Aboagye, E.O. and Derbyshire, H. (2000). “Gender Mainstreaming:

Emerging Lessons from Ghana”.

Alhassan, E. and Odame, F.S. (2015). “Gender Inequality in Basic Education in the

Northern Region of Ghana: Household and Contextual Factors in Perspectives”,

Ghana Journal of Development Studies 12: 125–141.

248

Amoah, J. (2007). “The World on Her Shoulders: The Rights of the Girl-Child in the Context

of Culture and Identity”, Essex Human Rights Review 4 (2): 1–23.

Amoakohene, M.I. (2004). “Violence Against Women in Ghana: A Look at Women’s

Perceptions and Review of Policy and Social Responses”, Social Science & Medicine

59 (11): 2373–2385.

Armstrong, A.K. (1998). Culture and Choice: Lessons from Survivors of Gender Violence

in Zimbabwe. Chapel Hill, NC: AKA.

Artz, L. (2014). “Fear or Failure: Why Victims of Domestic Violence Retract from the

Criminal Justice Process”, South African Crime Quarterly 37: 3–10.

Atinmo, M. (2001). “Sociocultural Implications of Wife Beating Among the Yoruba in

Ibadan City, Nigeria”. In: Oyekanmi, Felicia (ed.), Men, Women and Violence. Dakar:

Council for the Development of Social Science Research in Africa.

Banda, F. (1995). “Provision of Maintenance for Women and Children in Zimbabwe”,

Cardozo Women's LJ 2 (71).

Barker, G. and Ricardo, C. (2005). Young Men and the Construction of Masculinity in Sub-

Saharan Africa: Implications for HIV/AIDS, Conflict, and Violence. Washington, DC:

World Bank.

Bell, C.C. and Mattis, J. (2000). “The Importance of Cultural Competence in Ministering to

African American Victims of Domestic Violence”, Violence Against Women 6: 515–

532.

Belknap, J. (2014). The Invisible Woman: Gender, Crime, and Justice. Boston, MA: Cengage

Learning.

Bloch, F. and Rao, V. 2002. “Terror as a Bargaining Instrument: A Case Study of Dowry

Violence in Rural India”, American Economic Review 92: 1029–1043.

Bograd, M. (1999). “Strengthening Domestic Violence Theories: Intersections of Race,

Class, Sexual orientation, and Gender”, Journal of Marital and Family Therapy 25: 275–

289.

Bourdieu, P. (2001). “Structure, Habitus, Practice”. In: Erickson, P. and Murphy, I. (eds),

Readings for a History of Anthropological Theory. Peterborough, ON: Broadview Press.

Bobonis, G.J., Gonzalez-Brenes, M. and Castro, R. (2013). “Public Transfers and Domestic

Violence: The Roles of Private Information and Spousal Control”, American Economic

Journal: Economic Policy 5 (1): 179–205.

Bowman, C.G. (2003). “Domestic Violence: Does the African Context Demand a Different

Approach?”, International Journal of Law and Psychiatry 26 (5): 473–491.

249

Bowman, C.G. (2002). “Theories of Domestic Violence in the African Context”, Journal of

Gender, Social Policy & the Law 11 (2).

Bowman, C.G. and Schneider, E.M. (1998). “Feminist Legal Theory, Feminist Lawmaking,

and the Legal Profession”, Fordham L. Rev. 67: 249.

Brickell, K. (2015a). “Clouding the Judgment of Domestic Violence Law: Victim Blaming

by Institutional Stakeholders in Cambodia”, Journal of Interpersonal Violence, June

2015: 1–21.

Brickell, K. (2015b). “Towards Intimate Geographies of Peace? Local Reconciliation of

Domestic Violence in Cambodia”, Transactions of the Institute of British Geographers

40 (3): 321–33.

Bronfenbrenner, U. (1979). The Ecology of Human Development: Experiments by Nature and

Design. Cambridge, MA: Harvard University Press.

Burgess, R. and Campbell, C. (2015). “Creating Social Policy to Support Women's Agency

in Coercive Settings: A Case Study from Uganda”, Global Public Health 1–17.

Butegwa, F. (1994). “Using the African Charter on Human Rights and People’s Rights to

Secure Women's Access to Land in Africa”. In: R.J. Cook (ed.), Human Rights of

Women: National and International Perspectives. Philadelphia, PA: University of

Pennsylvania Press: 495.

Buzawa, E.S. and Buzawa, C.G. (2003). Domestic Violence: The Criminal Justice Response. New

York: Sage.

Campbell, C. (1997). “Migrancy, Masculine Identities and AIDS: The Psychosocial Context

of HIV Transmission on the South African Gold Mines”, Social Science & Medicine

45: 273–281.

Carlson, B.E. (1984). “Causes and Maintenance of Domestic Violence: An Ecological

Analysis”, Social Service Review 58: 569–587.

Carrell, S.E. and Hoekstra, M.L. (2010). “Externalities in the Classroom: How Children

Exposed to Domestic Violence Affect Everyone's Kids”, American Economic Journal:

Applied Economics 2 (1): 211–228.

Cerulli, C., Kothari, C., Dichter, M., Marcus, S., Kim, T.K., Wiley, J. and Rhodes, K.V. (2015).

“Help-Seeking Patterns among Women Experiencing Intimate Partner Violence: Do

They Forgo the Criminal Justice System if their Adjudication Wishes Are Not Met?”,

Violence and Victims 30: 16–31.

Clark, C.J., Silverman, J.G., Shahrouri, M., Everson-Rose, S. and Groce, N. (2010). “The Role

of the Extended Family on Women's Risk of Intimate Partner Violence in Jordan”,

Social Science and Medicine 70: 144–151.

250

Coker-Appiah, D. and Cusack, K. (1999). Breaking the Silence and Challenging the Myths of

Violence Against Women and Children in Ghana: Report of a National Study on Violence.

Accra-North, Ghana: Gender Studies & Human Rights Documentation Centre.

Cook, R.J. (2011). Human Rights of Women: National and International Perspectives.

Philadelphia, PA: University of Pennsylvania Press.

Cornwall, A., Harrison, E. and Whitehead, A. (2007). “Gender Myths and Feminist Fables:

The Struggle for Interpretive Power in Gender and Development”, Development and

Change 38: 1–20.

Cornwall, A. and White, S. (2000). “Men, Masculinities and Development: Politics, Policies

and Practices”, IDS Bulletin 31, 1–6.

Coughtry, S.E. (2011). “Patriarchy and the Trap of Masculinity: A Post-Colonial Analysis

of Violence against Sexual Minorities in Uganda”. MSW Thesis. Northampton, MA:

Smith College for Social Work.

Cusack, K. and Manuh, T. (2009) (eds.) The Architecture for Violence against Women in Ghana.

Accra-North, Ghana: Gender Studies & Human Rights Documentation Centre.

http://gendercentreghana.org/?p=632.

Daniel, W.W. (1999). Biostatistics: a foundation for analysis in the health sciences. Hoboken, NJ:

Wiley.

Deaton, A. (1997). The Analysis of Household Surveys: A Microeconometric Approach to

Development Policy. Baltimore, MD: Johns Hopkins University Press.

Dobash, R.E. and Dobash, R. (1979). Violence against Wives. New York: Free Press.

Dobash, R.E. and Dobash, R. (1977). “Love, Honour and Obey: Institutional Ideologies and

the Struggle for Battered Women”, Crime, Law and Social Change 1 (4): 403–415.

Doughty, K.C. (2014). “‘Our Goal Is Not to Punish but to Reconcile’: Mediation in

Postgenocide Rwanda”, American Anthropologist 116: 780–794.

Duvvury, N., Carney, P. and Minh, N. (2012). Estimating the Cost of Domestic Violence in Viet

Nam. New York: UN Women.

Dyson, T. and Moore, M. 1983. “Kinship Structure, Female Autonomy and Demographic

Behavior in India”, Population and Development Review 9: 35–60.

Edström, J., Das, A. and Dolan, C. (2014a). “Introduction: Undressing Patriarchy and

Masculinities to Re-politicise Gender”, IDS Bulletin 45: 1–10.

Edström, J., Izugbara, C., Nesbitt-Ahmed, Z., Otieno, P.E., Granvik, M. and Matindi, S.

(2014b). “Men in Collective Action on SGBV in Kenya: A Case Study”, IDS Evidence

Report 70. Brighton, UK: IDS.

http://gendercentreghana.org/?p=632

251

Ellsberg, M. and Heise, L. (2005). Researching Violence Against Women: A Practical Guide for

Researchers and Activists. Geneva: World Health Organization, PATH.

Ellsberg, M., Jansen, H.A., Heise, L., Watts, C.H. and Garcia-Moreno, C. (2008). “Intimate

Partner Violence and Women's Physical and Mental Health in the WHO Multi-

country Study on Women’s Health and Domestic Violence: An Observational

Study”, Lancet 371: 1165–1172.

Emmel, N. (2008). Participatory Mapping: An innovative sociological method (Toolkit No. 3).

Southampton, UK: ESRC National Centre for Research Methods.

Eswaran, M. and Malthotra, N. (2011). “Domestic Violence and Women's Autonomy in

Developing Countries: Theory and Evidence”, Canadian Journal of Economics 44 (4):

1222–1263.

Eves, R. (2010). “Masculinity Matters: Men, Gender-Based Violence and the AIDS

Epidemic in Papua New Guinea”, Civic Insecurity: Law, Order and HIV in Papua New

Guinea, 1.

Ewelukwa, U. (2002). “Post-Colonialism, Gender, Customary Injustice: Widows in African

Societies”, Human Rights Quarterly 24: 424–486.

Freedman, J. and Jacobson, S. (2012). “Masculinities, Gender, and Violence”, Engaging Men

in the Fight Against Gender Violence: Case Studies from Africa, 1.

Freeman, M.D. (1977). “Le Vice Anglais? Wife-Battering in English and American Law”,

Family Law Quarterly: 199–251.

Friedman, L. and Couper, S. (1987). The Cost of Domestic Violence: A Preliminary Investigation

of the Financial Cost of Domestic Violence. New York, Victims Services Agency.

Fulu, E., Kerr-Wilson, A., Lang, J., Gibbs, A., Jacobson, J., Jewkes, R., Mazars, C.,

Shauerhammer, V. and Watts, C. 2014. “What Works to Prevent Violence Against

Women and Girls: Evidence Review of Interventions to Prevent Violence Against

Women and Men”.

Gallegos, J.V. and Gutierrez, A. (2011). “The Effect of Civil Conflict on Domestic Violence:

The Case of Peru”, unpublished manuscript.

Garcia-Moreno, C., Jansen, H.A., Ellsberg, M., Heise, L. and Watts, C.H. (2006). “Prevalence

of Intimate Partner Violence: Findings from the WHO Multi-country Study on

Women's Health and Domestic Violence”, Lancet 368: 1260–1269.

García-Moreno, C., Jansen, H.A. Ellsberg, M., Heise, L. and Watts, C.H. (2005). WHO Multi-

country Study on Women’s Health and Domestic Violence Against Women: Initial Results

on Prevalence, Health Outcomes and Women’s Responses. Geneva: World Health

Organization.

252

García-Moreno, C., World Health Organization, London School of Hygiene and Tropical

Medicine and South African Medical Research Council (2013). Global and Regional

Estimates of Violence Against Women: Prevalence and Health Effects of Intimate Partner

Violence and Non-Partner Sexual Violence. Geneva: World Health Organization.

Ghana Statistical Service (GSS) (2013). 2010 Population and Housing Census: National

Analytical Report. Accra: Ghana Statistical Service.

Ghana Statistical Service, Ghana Health Service and ICF Macro (2009). Ghana Demographic

and Health Survey 2008 (DHS Final Reports No. FR221). Accra: Ghana Statistical

Service and Ghana Health Service, and Calverton, MD: ICF Macro.

Geisler, G. (1995). “Troubled Sisterhood: Women and Politics in Southern Africa: Case

Studies from Zambia, Zimbabwe and Botswana”, African Affairs 545–578.

Glaser, B.G. and Strauss, A. (1998). “Grounded Theory”, Strategien qualitativer Forschung.

Bern: Huber.

Goode, W.J. (1971). “Force and Violence in the Family”, Journal of Marriage and the Family

33: 624–636.

Gosselin, D.K. (2000). Heavy hands: An introduction to the crimes of domestic violence. Upper

Saddle River, NJ: Prentice Hall.

Gupte, J., Justino, P. and Tranchant, J.-P. (2014). “Households amid Urban Riots: The

Economic Consequences of Civil Violence in India”, Journal of Conflict Resolution 58

(8): 1445–1473.

Haraway, D. (1988). “Situated Knowledges: The Science Question in Feminism and the

Privilege of Partial Perspective”, Feminist Studies 575–599.

Harrison, A., O’Sullivan, L., Hoffman, S., Dolezal, C. and Morrell, R. (2006). “Gender Role

and Relationship Norms among Young Adults in South Africa: Measuring the

Context of Masculinity and HIV Risk”, Journal of Urban Health 83: 709–722.

Heath, R. (2014). “Women’s Access to Labor Market Oportunities, Control of Household

Resources, and Domestic Violence: Evidence from Bangladesh”, World Development

57 (C): 32–46.

Heise, L.L. (1998). “Violence against Women An Integrated, Ecological Framework”,

Violence Against Women 4(3): 262–290.

Heise, L.L. and Garcia-Moreno, C. (2002). “Violence by Intimate Partners”. In: E.G. King,

L.L. Dahlberg, J.A. Mercy, A.B. Zwi and R. Lozano (eds.), World Report on Violence

and Health. Geneva: World Health Organization.

Hidrobo, M. and Fernald, L. (2013). “Cash Transfers and Domestic Violence”, Journal of

Health Economics 32: 304–319.

253

Hindin, M.J., Kishor, S. and Ansara, D.L. (2008). “Intimate Partner Violence Among

Couples in 10 DHS Countries: Predictors And Health Outcomes”’, DHS Analytical

Studies No. 18. Calverton, MD: Macro International Inc.

Hodgson, D.L. (2002). “Women's Rights as Human Rights: Women in Law and

Development in Africa (WILDAF)”, Africa Today 49: 3–26.

Hoyle, C. (1998). Negotiating Domestic Violence: Police, Criminal Justice, and Victims. Oxford,

UK: Oxford University Press.

Hunnicutt, G. (2009). “Varieties of Patriarchy and Violence against Women: 'Resurrecting

Patriarchy' as a Theoretical Tool”, Violence against Women 15, (5): 553–573.

IDS, Promundo and Sonke Gender Justice (2015). “Engendering Men: A Collaborative

Review of Evidence on Men and Boys in Social Change and Gender Equality”,

EMERGE Evidence Review Report. Brighton, UK: IDS.

Jewkes, R. (2002). “Intimate Partner Violence: Causes and Prevention”, Lancet 359 (9315):

1423–1429.

Jewkes, R., Levin, J. and Penn-Kekana, L. (2002). “Risk Factors for Domestic Violence:

Findings from a South African Cross-Sectional Study”, Social Science & Medicine 55

(9): 1603–1617.

Jones, A. (2006). “Straight as a Rule Heteronormativity, Gendercide, and the

Noncombatant Male”, Men and Masculinities 8: 451–469.

Jones, R.B. (2011). “Development of Feminist Postcolonial Theory”, Postcolonial

Representations of Women. Dordrecht: Springer Netherlands.

Justino, P., Leone, M. and Salardi, P. (2015). “Does War Empower Women? Evidence from

Timor Leste”, IDS Evidence Report 121. Brighton, UK: IDS.

Justino, P., Cardona, I., Mitchell, B. and Muller, C. (2012). “The Role of Women in Post-

Conflict Economic Recovery”. New York: UN Women.

Kelly, K.A. (2003). Domestic violence and the politics of privacy. Ithaca, NY: Cornell University

Press.

Kimuna, S.R. and Djamba, Y.K. (2008). “Gender Based Violence: Correlates of Physical and

Sexual Wife Abuse in Kenya”, Journal of Family Violence 23: 333–342.

Kishor, S. and Bradley, S.E.K. (2012). “Women’s and Men’s Experience of Spousal Violence

in Two African Countries: Does Gender Matter?”, DHS Analytical Studies No. 27.

Calverton, MD: ICF International.

Koenen, K., Moffitt, T., Caspi, A., Taylor, A. and Purcell, S. (2003). “Domestic Violence is

Associated with Environmental Suppression of IQ in Young Children”, Development

and Psychopathology 15 (2): 297–311.

254

Koenig, M.A., Lutalo, T., Zhao, F., Nalugoda, F., Wabwire-Mangen, F., Kiwanuka, N.,

Wagman, J., Serwadda, D., Wawer, M. and Gray, R. (2003). “Domestic Violence in

Rural Uganda: Evidence from a Community-Based Study”, Bulletin of the World

Health Organization 81: 53–60.

La Mattina, G. (2013). Armed Conflict and Domestic Violence: Evidence from Rwanda. Technical

Report. Boston, MA: Boston University.

Little, J. and Jones, O. (2000). “Masculinity, Gender, and Rural Policy”, Rural Sociology 65:

621.

Lloyd, S. (1997). “The Effects of Domestic Violence on Women's Employment”, Law &

Politics 19 (2): 139–167.

Lugo, B.L. (2015). An Exploratory Study on Mental Health Effects of Therapist Minimization of

Domestic Violence Victims’ Experiences. San Diego, CA: Alliant International

University.

Maguigan, H. (2002). “Wading into Professor Schneider’s Murky Middle Ground between

Acceptance and Rejection of Criminal Justice Responses to Domestic Violence”, Am.

UJ Gender Soc. Pol’y & L. 11: 427.

Mann, J.R. and Takyi, B.K. (2009). “Autonomy, Dependence or Culture: Examining the

Impact of Resources and Socio-Cultural Processes on Attitudes Towards Intimate

Partner Violence in Ghana, Africa”, Journal of Family Violence 24 (5): 323–335.

Mann, R.M. (2000). Who Owns Domestic Abuse?: The Local Politics of a Social Problem. Toronto,

ON: University of Toronto Press.

Manuh,T. (2007). “The Passage of Domestic Violence Legislation in Ghana”,

http://r4d.dfid.gov.uk/PDF/Outputs/WomenEmp/GhanaDV.pdf.

Martin, D. (1976). Battered Wives. San Francisco, CA: Glide Publications.

McFadden, P. (2000). “The State of Feminism in Africa Today”, News from the Nordic Africa

Institute 2.

Menon, N. and Johnson, M.P. 2007. “Patriarchy and Paternalism in Intimate Partner

Violence: A Study of Domestic Violence in Rural India”. In: K.K. Misra and J.H.

Lowry (eds), Recent Studies on Indian Women: Empirical Work of Social Scientists.

Jaipur: Rawat Publications.

Mills, E., De Paoli, M.M. and Grønningsaeter, A. (2009). “Love in the Time of AIDS: The

Relational Gender Dynamics of Prevention, Testing and Treatment”, CSSR Working

Paper Series 252. Cape Towm, South Africa: Centre for Social Science Research.

http://r4d.dfid.gov.uk/PDF/Outputs/WomenEmp/GhanaDV.pdf

255

Mills, E., Nesbitt-Ahmed, Z., Diggins, J. and Mackieu, T.D. (2015a). ‘They Call Me Warrior’:

The Legacy of Conflict and the Struggle to End Sexual and Gender-based Violence in Sierra

Leone. Brighton, UK: IDS.

Mills, E., Saharokh, T., Wheeler, J., Black, G., Cornelius, R. and Van Den Heever, L. (2015b).

Turning the Tide: The Role of Collective Action for Addressing Structural and Gender-based

Violence in South Africa. Brighton, UK: IDS.

Minayo, M.C.D.S. (1994). “Inequality, Violence, and Ecology in Brazil”, Cadernos de Saúde

Pública 10: 241–250.

Mitchell, L. (2011). Service Users’ Perceptions of the Domestic Violence and Victims’ Support

Unit, Ghana Police Service. Accra: Voluntary Service Overseas (VSO) and Domestic

ViolenceSecretariat.

Moosa, Z. (2012). Violence Against Women and Girls in the Post-2015 Framework: Why & How.

London: ActionAid.

Murphy, J.C. (2002). “Engaging with the State: The Growing Reliance on Lawyers and

Judges to Protect Battered Women”, Am. UJ Gender Soc. Pol’y & L. 11: 499.

NETRIGHT (2012). “Ghana NGOs Alternative Report for CSW 57,” Accra-North:

NETRIGHT, Association of African Women in Research & Development

(AAWORD) and UN Women Ghana Country Programme Office.

Njovana, E. and Watts, C. (1996). “Gender Violence in Zimbabwe: A Need for

Collaborative Action”, Reproductive Health Matters 4: 46–55.

Noe, D. and Rieckmann, J. (2013). “Violent Behaviour: The Effect of Civil Conflict on

Domestic Violence in Colombia”, Poverty, Equity and Growth Discussion Paper.

Göttingen, Germany: Courant Research Centre.

Ofei-Aboagye, R.O. (1994a). “Domestic Violence in Ghana: An Initial Step”, Colum. J.

Gender & L., 4: 1.

Ofei-Aboagye, R.O. (1994b). “Initial Questions”, The Public Nature of Private Violence: The

Discovery of Domestic Abuse. New York: Routledge: 260.

Ofei-Aboagye, R.O. (1994c). “Altering the Strands of the Fabric: A Preliminary Look at

Domestic Violence in Ghana”, Signs 19 (4): 924–938.

Panda, P. and Agarwal, B. (2005). “Marital Violence, Human Development and Women's

Property Status in India”, World Development 33: 823–850.

Perilla, J.L., Bakeman, R. and Norris, F.H. (1994). “Culture and Domestic Violence: The

Ecology of Abused Latinas”, Violence and Victims 9: 325–339.

Pollack, R.A. (2004). “An Intergenerational Model of Domestic Violence”, Journal of

Population Economics 17: 311–329.

256

Pool, M.S., Otupiri, E., Owusu-Dabo, E., de Jonge, A. and Agyemang, C. (2014). “Physical

Violence during Pregnancy and Pregnancy Outcomes in Ghana”, BMC Pregnancy

and Childbirth 14 (1): 71.

Ribero, R. and Sanchez, F. (2005). “Determinants, Effects and Costs of Domestic Violence”,

Documento CEDE 2005-38. Bogota, Colombia: Universidad de los Andes.

Richards, P. (2005). “To Fight or to Farm? Agrarian Dimensions of the Mano River Conflicts

(Liberia and Sierra Leone)”, African Affairs 104: 571–590.

Richards, P. (1996). Fighting for the Rain Forest: War, Youth and Resources in Sierra Leone.

Oxford: James Currey.

Rowe, M. (2015). “Community Impact Statements in Domestic Violence Criminal Cases: A

World First”, Legaldate 27: 11.

Schneider, E.M. (2000). Battered Women and Feminist Lawmaking. New Haven, CT: Yale

University Press.

Schneider, E.M. (2008). “Domestic Violence Law Reform in the Twenty-First Century:

Looking Back and Looking Forward”, Fam. LQ 42: 353.

Shanley, M.L. (1993). Feminism, Marriage, and the Law in Victorian England. Princeton, NJ:

Princeton University Press.

Shepard, M. and Pence, E. (1988). “The Effect of Battering in the Employment Status of

Women”, Affilia 3 (2): 55–61.

Sherman, L.W., Schmidt, J.D. and Rogan, D.P. (1992). Policing Domestic Violence: Experiments

and Dilemmas. New York: Free Press.

Sipsma, H., Ofori-Atta, A., Canavan, M., Osei-Akoto, I., Udry, C. and Bradley, E. (2013).

“Poor Mental Health in Ghana: Who Is at Risk?”, BMC Public Health 13 (1): 288.

Sokoloff, N.J. and Dupont, I. (2005). “Domestic Violence at the Intersections of Race, Class,

and Gender Challenges and Contributions to Understanding Violence Against

Marginalized Women in Diverse Communities”, Violence against Women 11: 38–64.

Stockl, H., Devries, K., Rotstein, A., Abrahams, N., Campbell, J., Watts, C. and Garcia-

Moreno, C. (2013). “The Prevalence of Intimate Partner Homicide: A Systematic

Review”, Lancet 382.

Sternberg, K.J., Lamb, M.E., Greenbaum, C., Cicchetti, D., Dawud, S., Cortes, R., Krispin,

O. and Lorey, F. (1993). “Effects of Domestic Violence on Children's Behavior

Problems and Depression”, Developmental Psychology 29 (1): 44–52.

Sulak, T.N., Saxon, T.F. and Fearon, D. (2014). “Applying the Theory of Reasoned Action

to Domestic Violence Reporting Behavior: The Role of Sex and Victimization”,

Journal of Family Violence 29: 165–173.

257

Takyi, B.K. and Mann, J. (2006). “Intimate Partner Violence in Ghana, Africa: The

Perspectives of Men regarding Wife Beating”, International Journal of Sociology of the

Family 32 (1): 61–78.

Tamale, S.R. (1993). “Law Reform and Women's Rights in Uganda”, East African Journal of

Peace and Human Rights 1 (2): 164–94.

Tanchen, H.V., Witte, A.D. and Long, S.K. (1991). “Domestic Violence: A Nonrandom

Affair”, International Economic Review 32: 491–511.

Tenkorang, E.Y., Owusu, A.Y., Yeboah, E.H. and Bannerman, R. (2013). “Factors

Influencing Domestic and Marital Violence against Women in Ghana”, Journal of

Family Violence 28 (8): 771–81.

Thaler, K. (2012). “Norms About Intimate Partner Violence among Urban South Africans:

A Quantitative and Qualitative Vignette Analysis”, CSSR Working Paper No. 302.

Cape Towm, South Africa: Centre for Social Science Research, Social Surveys Unit.

Tsikata, D. and Seini, W. (2004). “Identities, Inequalities and Conflicts in Ghana”, CRISE

Working Paper 5. Oxford: Centre for Research on Inequality, Human Security and

Ethnicity.

Tsikata, P.Y. (2014). “The Historical and Contemporary Representation of Africa in Global

Media Flows: Can the Continent Speak Back for Itself on Its Own Terms?”,

Communication 40: 34-48.

Turshen, M. and Turshen, M. (2000). “Even Strong Backs Need Support: Introductory

Notes”, African Women’s Health 8.

UNITAR (2011). “River of Life Method”, www.click4it.org.

USAID (2006). Addressing Gender-based Violence through USAID’s Health Programs: A Guide

for Health Sector Program Officers. Washington, DC: USAID.

Walker, L.E. (1999). “Psychology and Domestic Violence Around the World”, American

Psychologist 54: 21.

Watts, C., Osam, S. and Win, E. (1995). “The Private is Public: A Study of Violence against

Women in Southern Africa”. Harare: Zimbabwe: Women in Law and Development

in Africa.

Wolfe, D.A., Crooks, C.V., Lee, V., McIntyre-Smith, A., Jaffe, P.G. (2003), “The Effects of

Children's Exposure to Domestic Violence: A Meta-Analysis and Critique”, Clinical

Child and family Psychology Review 6 (3): 171–187.

WHO (2001). “Putting Women First: Ethical and Safety Recommendations for Research on

Domestic Violence against Women”. Geneva: World Health Organization,

http://www.who.int/gender/documents/violence/who_fch_gwh_01.1/en/.

http://www.who.int/gender/documents/violence/who_fch_gwh_01.1/en/

258

Wilson, M. and Daly, M. (1996). “Male Sexual Proprietariness and Violence against Wives”,

Current Directions in Psychological Science 5: 2–7.

Wilson, M. and Daly, M. (1993). “An Evolutionary Psychological Perspective on Male

Sexual Proprietariness and Violence against Wives”, Violence and Victims 8: 271–294.

Yllo, K.A. and Strauss, M.A. (1990). “Patriarchy and Violence against Wives: The Impact of

Structural and Normative Factors”. In: M.A. Strauss and R.J. Gelles (eds), Physical

Violence in American Families. New Brunswick, NJ: Transaction Publishers.

Yodanis, C.L. (2004). “Gender Inequality, Violence against Women, and Fear A Cross-

National Test of the Feminist Theory of Violence Against Women”, Journal of

Interpersonal Violence 19: 655–675.

259

Appendices

260

Appendix A: Incidence of domestic violence worldwide

Prevalence of physical

violence against women (%)

Prevalence of sexual violence

against women (%)

All

perpetrators

By intimate

partner

All

perpetrators

Intimate

partner

Region Country or area Year Life

time

Last

12

mths

Life

time

Last

12

mths

Life

time

Last

12

mths

Life

time

Last

12

mths

Africa Burkina Faso 2010 19.8 9.4 11.1 9.2 .. .. 1.5 1.2

Africa Cabo Verde 2005 21.5 20.3 15.7 .. .. .. 3.6 ..

Africa Cameroon 2011 54.6 27.4 44.8 29.0 29.0 9.8 20.3 11.2

Africa Central African

Republic

2006 35.3 28.7 25.4 22.6 .. .. 11.6 8.3

Africa Chad 2010 .. .. .. 18.4 .. .. .. 12.0

Africa Comoros 2012 14.0 5.8 5.6 4.2 6.0 1.3 1.8 1.3

Africa Côte d'Ivoire 2011-12 35.6 19.9 24.6 22.2 4.5 .. 5.3 4.6

Africa Democratic

Republic of the

Congo

2007 63.7 49.0 56.9 .. 16.0 4.2 35.3 ..

Africa Egypt 2005 47.4 15.7 33.2 18.2 .. .. 6.6 3.9

Africa Equatorial Guinea 2011 62.8 55.8 54.4 40.2 31.9 8.2 17.4 12.5

Africa Ethiopia -

Province

2002 .. .. 48.7 29.0 .. .. 58.6 44.4

Africa Gabon 2012 52.4 22.3 46.2 28.3 20.8 8.3 17.0 11.8

Africa Ghana 2008 36.6 17.2 20.6 18.0 18.8 .. 8.2 5.2

Africa Kenya 2008-09 38.5 24.0 37.0 31.3 20.6 .. 17.2 13.6

Africa Liberia 2007 44.0 28.9 35.0 33.0 17.6 .. 10.8 9.6

Africa Malawi 2010 28.2 14.2 21.7 14.7 25.3 .. 18.9 13.4

Africa Mali 2012-13 38.3 24.7 29.8 20.7 12.6 10.5 13.9 12.1

Africa Morocco 2009-10 35.3 15.2 .. 6.4 22.6 8.7 .. 6.6

Africa Mozambique 2011 33.4 25.0 31.5 25.9 12.3 6.9 7.9 6.9

Africa Namibia – City 2002 .. .. 30.6 15.9 .. .. 16.5 9.1

Africa Nigeria 2013 27.8 11.2 14.4 9.3 7.4 3.3 4.8 3.7

Africa Rwanda 2010 41.2 .. 55.6 32.8 22.3 .. 17.5 13.3

Africa Sao Tome and

Principe

2008-09 33.4 21.1 26.5 .. 13.1 .. 8.3 ..

Africa Sierra Leone 2013 55.5 27.0 44.2 27.2 10.5 4.9 7.3 5.1

Africa South Africa 1998 .. .. 12.5 6.3 .. .. 4.4 ..

Africa Tunisia 2010 31.7 7.3 20.3 7.2 15.7 7.4 14.2 9.0

Africa Uganda 2011 56.1 26.9 42.7 24.9 27.8 16.2 27.3 20.9

Africa United Republic

of Tanzania

2010 38.7 33.0 39.2 33.4 20.3 .. 17.2 13.7

Africa Zambia 2007 46.8 32.5 46.5 39.6 20.2 .. 16.7 16.0

Africa Zimbabwe 2010-11 29.9 18.4 28.8 20.7 27.2 .. 26.0 13.3

Asia Armenia 2008 .. .. 8.9 .. .. .. 3.3 ..

Asia Azerbaijan 2006 13.3 7.7 12.8 9.7 3.7 .. 2.9 2.0

Asia Bangladesh 2011 .. .. 64.6 46.4 .. .. 36.5 24.1

261

Asia Cambodia 2005 22.3 10.3 12.8 7.9 .. .. 2.7 1.7

Asia China 1999-00 .. .. 15.4 .. .. .. .. ..

Asia China, Hong

Kong

2005 12.0 2.0 6.0 1.0 14.0 3.0 5.0 1.0

Asia Cyprus 2014 20.0 5.0 14.0 3.0 5.0 1.0 4.0 1.0

Asia Georgia 2010 .. .. 4.5 1.4 .. .. 1.7 0.5

Asia India 2005-06 33.5 18.9 35.1 21.4 8.5 .. 10.0 7.2

Asia Japan 2010 .. .. 25.9 .. .. .. 14.1 ..

Asia Jordan 2012 34.3 12.6 21.1 11.2 9.2 .. 8.6 6.0

Asia Kyrgyzstan 2012 23.1 13.3 25.1 16.9 3.4 2.1 4.0 2.8

Asia Maldives 2006 .. .. 17.9 5.7 .. .. 6.7 2.0

Asia Mongolia 2008 .. .. .. .. 4.0 .. .. ..

Asia Nepal 2011 21.5 9.3 23.1 10.4 12.3 6.4 14.3 7.7

Asia Pakistan 2012-13 32.2 19.2 26.8 18.0 .. .. .. ..

Asia Philippines 2013 19.6 5.6 12.7 5.3 6.3 2.7 5.3 3.2

Asia Republic of Korea 2013 .. .. .. 7.2 19.5 2.7 .. 5.4

Asia Singapore 2009 6.8 1.0 5.7 0.9 4.2 0.3 1.2 0.1

Asia State of Palestine 2011 .. .. 30.6 23.5 .. .. 14.6 11.8

Asia Tajikistan 2012 18.8 13.0 19.5 14.5 3.7 2.5 4.4 3.3

Asia Thailand -

Province

2002 .. .. 33.8 13.4 .. .. 28.9 15.6

Asia Thailand - City 2005 .. .. 22.9 7.9 .. .. 29.9 17.1

Asia Timor-Leste 2009-10 38.1 29.2 33.5 30.7 3.4 .. 2.9 2.0

Asia Turkey 2014 .. .. 36.0 8.0 .. .. 12.0 5.0

Asia Viet Nam 2010 35.2 .. 31.5 6.4 10.8 .. 9.9 4.2

Europe Albania 2013 .. .. 23.7 14.7 .. .. 7.9 5

Europe Austria 2014 17.0 4.0 12.0 2.0 9.0 2.0 6.0 1.0

Europe Belgium 2014 33.0 10.0 22.0 5.0 13.0 2.0 9.0 1.0

Europe Bulgaria 2014 27.0 7.0 22.0 6.0 12.0 3.0 9.0 3.0

Europe Croatia 2014 19.0 5.0 12.0 3.0 5.0 1.0 3.0 0.0

Europe Czech Republic 2014 30.0 7.0 19.0 4.0 9.0 2.0 7.0 1.0

Europe Denmark 2014 48.0 10.0 29.0 3.0 19.0 2.0 11.0 1.0

Europe Estonia 2014 30.0 4.0 19.0 2.0 13.0 2.0 7.0 1.0

Europe Finland 2014 43.0 9.0 27.0 4.0 17.0 3.0 11.0 1.0

Europe France 2014 42.0 11.0 25.0 5.0 15.0 2.0 9.0 1.0

Europe Germany 2014 33.0 7.0 20.0 3.0 12.0 1.0 8.0 1.0

Europe Greece 2014 24.0 7.0 18.0 5.0 6.0 2.0 5.0 2.0

Europe Hungary 2014 25.0 8.0 19.0 5.0 9.0 3.0 7.0 2.0

Europe Iceland 2008 29.8 2.1 19.7 1.2 24.2 1.6 6.1 0.1

Europe Ireland 2014 24.0 7.0 14.0 3.0 8.0 2.0 6.0 1.0

Europe Italy 2014 25.0 6.0 17.0 5.0 9.0 4.0 7.0 4.0

Europe Latvia 2014 36.0 6.0 31.0 5.0 14.0 2.0 9.0 1.0

Europe Lithuania 2014 30.0 5.0 24.0 4.0 7.0 0.0 4.0 0.0

Europe Luxembourg 2014 36.0 6.0 21.0 3.0 15.0 2.0 9.0 1.0

Europe Malta 2014 19.0 4.0 13.0 3.0 9.0 2.0 6.0 1.0

Europe Netherlands 2014 41.0 9.0 22.0 4.0 18.0 3.0 11.0 2.0

262

Europe Norway 2008 .. .. 13.9 .. .. .. 9.4 1.0

Europe Poland 2014 18.0 4.0 12.0 2.0 5.0 1.0 4.0 1.0

Europe Portugal 2014 23.0 5.0 18.0 4.0 4.0 1.0 3.0 1.0

Europe Republic of

Moldova

2010 .. .. 39.7 8.9 .. .. 18.6 4.1

Europe Romania 2014 29.0 7.0 23.0 6.0 6.0 2.0 5.0 2.0

Europe Serbia - City 2003 .. .. 22.8 3.2 .. .. 6.3 1.1

Europe Slovakia 2014 33.0 9.0 22.0 6.0 10.0 3.0 8.0 2.0

Europe Slovenia 2014 21.0 3.0 12.0 2.0 7.0 1.0 4.0 0.0

Europe Spain 2014 20.0 3.0 12.0 1.0 6.0 1.0 4.0 1.0

Europe Sweden 2014 41.0 8.0 24.0 3.0 18.0 3.0 10.0 2.0

Europe Switzerland 2003 27.0 1.0 9.0 1.0 25.0 1.0 3.0 ..

Europe Ukraine 2007 16.9 8.8 12.7 10.4 5.4 .. 3.3 2.2

Europe United Kingdom 2014 42.0 8.0 28.0 4.0 14.0 2.0 10.0 1.0

LA & Carib. Bolivia 2008 .. .. 23.1 .. .. .. 6.4 ..

LA & Carib. Brazil - City 2000-03 .. .. 27.2 8.3 .. .. 10.1 2.8

LA & Carib. Brazil - Province 2000-03 .. .. 33.8 12.9 .. .. 14.3 5.6

LA & Carib. Colombia 2010 .. .. 37.4 .. .. .. 9.7 ..

LA & Carib. Costa Rica 2003 47.0 11.0 33.0 7.0 41.0 7.0 15.0 3.0

LA & Carib. Dominican

Republic

2007 19.7 9.7 16.1 10.9 10.1 .. 5.5 3.6

LA & Carib. Ecuador 2011 38.0 .. 35.0 .. 25.7 .. 14.5 ..

LA & Carib. El Salvador 2008 .. .. 24.2 6.8 13.4 .. 11.5 3.3

LA & Carib. Guatemala 2008-09 .. .. 24.5 7.8 5.8 .. 12.3 4.8

LA & Carib. Haiti 2012 28.2 10.4 15.6 10.3 13.1 .. 11.1 8.6

LA & Carib. Honduras 2011-12 27.0 .. .. 10.0 4.0 .. .. 3.2

LA & Carib. Jamaica 2008-09 .. .. 17.0 6.5 12.0 .. 7.6 2.8

LA & Carib. Mexico 2011 15.2 6.4 12.3 5.4 38.9 20.8 5.9 2.5

LA & Carib. Nicaragua 2006-07 19.2 .. 27.0 8.0 11.4 .. 13.1 4.4

LA & Carib. Paraguay 2008 .. .. 17.9 6.7 .. .. 5.0 1.7

LA & Carib. Peru 2013 .. .. 35.7 11.5 .. .. 8.4 3.0

LA & Carib. Puerto Rico 1995-96 .. .. .. .. 9.1 .. 5.7 ..

Northern

America

Canada 2009 .. 3.4 .. 1.3 .. 2.0 .. ..

Northern

America

United States of

America

2011 .. .. 31.5 4.0 19.3 1.6 8.8 0.8

Oceania Australia 2012 34.0 4.6 15.6 .. 19.0 1.2 4.9 ..

Oceania Fiji 2010-11 68.5 .. 61.1 19.4 35.6 .. 33.9 14.2

Oceania Kiribati 2008 .. .. 60.0 32.4 .. .. 46.4 33.7

Oceania New Zealand -

City

2002 .. .. 30.2 5.3 .. .. 14.1 2.1

Oceania Samoa 2000 .. .. 40.5 17.9 .. .. 19.5 11.5

Oceania Solomon Islands 2009 .. .. 45.5 .. .. .. 54.7 ..

Oceania Tonga 2009 76.8 .. 33.4 12.5 17.4 .. 16.5 11.0

Oceania Vanuatu 2009 .. .. 51.0 33.0 .. .. 44.0 33.0

263

Appendix B: Timeline of major policies related to domestic and gender-based

violence International Regional National

1948•Universal Declaration of Human Rights

1966•International Covenant on Civil and Political Rights•International Covenant on Social, Economic and Cultural Rights

1981•African Charter on Human and Peoples' Rights

1982•Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW)

1992•Republic of Ghana Constitution - does not include domestic violence or violence against women

but does include equality for women

1993•Declaration on the Elimination of Violence Against Women - comprehensive set of standards in

international law for the protection of women against sexual and gender-based violence

1994•Republic of Ghana Criminal Code Article 69A - makes female genital mutilation a second-degree

felony

1994•African Platform for Action on Women - Dakar Declaration - "violation of women's rights is

violation of human rights" … appointment of a Special Rapporteur on Violence Against Women

1995•Protocol to the African Charter on Human and People's Rights on the Rights of Women in Africa

- includes the need for States to adopt legislation aimed to eradicate violence against women

1995

•Beijing Declaration and Platform for Action - Conference Report from Fourth World Conference on Women - Countries (including Ghana) that have signed up to the Beijing Declaration are requried to give five-yearly reports detailing the action they have taken to improve the situation for women and girls

1999

•African Commission on Human and People's Rights recruits a Special Rapporteur on the Rights of Women with a mandate to servce as a focal point for promotion and protection of the rights of women

2000•UN Resolution 55/68 - Elimination fo all forms of violence, including crimes against women

264

2003•UN Resolution 58/185 - In-depth study on all forms of violence against women

2003•Ghanian Poverty Reduction Strategy (GPRS) - includes the necessity to formulate and enact

Domestic Violence Act. Put together in consultation with the International Monetary Fund

2003

•Ghanaian Strategic Implementation Plan - includes 'mainstreaming of gender issues', decentralisation of 'women's machinery' and monitoring of policies affecting women and children

2004•UN Resolution 18/147 Elimination of Domestic Violence Against Women - specifically

addressing domestic violence as a separate issue for the first time

2004•African Union - Solemn Declaration on Gender Equality and Women's Empowerment - no

specific mention of domestic violence

2006•UN Resolution 61/143 - intensification of efforts to eliminate all forms of violence against

women

2007•African Commission on Human and People's Rights - Resolution 111 - On the right to remedy

and reparation for women and girl victims of sexual violence

2007•Republic of Ghana Domestic Violence Act

2008•UN Resolution 7/24 - Elimination of Violence Against Women

2009•UN Resolution 11/2 - Accelerating efforts to eliminate all forms of violence against women•UN Resolution 12/17 - Elimination of discrimination against women

2010

•GHANAP 1325 - the Ghanaian Action Plan for the implementation of UN Resolution 1325 (women, peace, security). Although not explicitly about domestic violence, the mechanisms that have been establised to protect women from domestic violence will be adopted to protect women in situations of conflict and gender-based violence

2010

•Ghanaian Sector Medium-Term Development Plan (SMTDP)- includes the need to education women regarding their human rights and access to the law - for example, the Domestic Violence Act

2010•Ghanaian Shared Growth and Development Agenda (GSGDA) - the Domestic Violence Act is seen

as the solution to women's inequality

265

Appendix C: Summary of qualitative sample, methods and instruments

Method of data

collection

Sampling Sample characteristics Instruments

FOCUS GROUP

DISCUSSIONS

One urban and one

rural site from one

district in each

region

1 hour /8–12 people

80 FGDs: 8 per region (4 in

rural, 4 in urban sites)

Group 1: Married/living

together men (aged 18–50)

Group 2: Married /living

together women (aged 18–50)

Group 3: Unmarried/youth

(aged 18–25)

Group 4: Mixed (aged 50+)

Semi-structured

questionnaire

OBSERVATIONS

AND IN-DEPTH

INTERVIEWS

Primary public

institutions of

victim support:

police and health in

one urban and one

rural site from one

district in each

region

Interviews with 20 DOVVSU

officers (2 from each chosen

district office in each region)

Observations conducted in

each of the 10 district

DOVVSUs visited where staff

were interviewed

28 health staff interviewed,

including doctors, nurses and

hospital administrators; 10

district hospitals were

observed

Observations

using a checklist

with additional

descriptions

Semi-structured

questionnaire

IN-DEPTH

INTERVIEWS

with OPINION

LEADERS,

ACTIVISTS and

LEGAL EXPERTS

Purposive sample

in all regions

equally divided

across urban and

rural areas

Total of 200 opinion leaders:34

10 legal practitioners (one per

region); 20 activists (2 per

region); 20 media persons (2

per region); 102 community

leaders as described in the

text

Observations

using a checklist

with additional

descriptions

Semi-structured

questionnaire

34 Legal practitioners, media and activists were mainly located at urban sites.

266

Appendix D: Distribution of enumeration areas across regions and survey

weights

Table D1: Distribution of enumeration areas

Region

2010

population

in household

distribution

2010

population

in household

distribution

per cent

Propor-

tionate

allocation

of

selected

EAs

Adjusted

number of

HHs

selected

per EA

Propor-

tionate

allocation

of

selected

EAs

Adjusted

number of

HHs

selected

per EA

Western 2,307,395 9.6 32 480 32 480

Central 2,113,766 8.9 29 435 29 435

Greater Accra 3,888,512 16.3 53 795 53 795

Volta 2,086,567 8.6 29 435 29 435

Eastern 2,574,549 10.7 35 525 35 525

Ashanti 4,671,982 19.4 64 960 64 960

Brong Ahafo 2,265,458 9.4 31 465 31 465

Northern 2,445,061 10.1 34 510 34 510

Upper East 1,034,704 4.2 14 210 14 210

Upper West 688,333 2.8 9 135 9 135

Total 24,076,327 100.0 331 4,950 331 4,950


Weighting of data

The survey data were weighted at the data management stage, taking into account the

probability of selecting a primary sampling unit from a region, the probability of selecting

a household from the primary sampling unit, and the probability of selecting a qualified

individual for interviewing. The weight was standardised to be applicable to the data. This

was because the sampling design was not self-weighting: households were selected with

unequal probabilities at the second stage of sampling. As a result, a sampling weight

accounted for differences in the selection probability and in the number of completed

interviews among primary sampling units. The required information for computing the

weights at the level of the primary sampling unit and household selection stage are

explained below.

Computation of weights

The weight is the product of the first- and second-stage probabilities of selection. The first-

stage probability of selection was calculated based on the number of clusters selected in

the stratum (ah), multiplied by household size (Mhi) (among those aged 15–60 years) and

divided by stratum size (∑Mhi). The second-stage probability of selection was computed as

the number of households selected in the clusters i (b) divided by the number of

households listed in cluster i (Mhi*) (i.e. the probability of selecting 15 households aged 15–

267

60 years out of the listed households aged 15–60 years). The reciprocal of the overall

probability of selection is the weighting factor, also known as the design weight.

Let Mhi = Number of 2010 Population Census households in the ith selected

enumeration area in the hth stratum or region

Mhi* = Number of households listed in the ith selected enumeration area in the hth

stratum (urban/rural in the region)

ΣMhi = Total number of households in the ith stratum (i.e. number of households

in

either an urban or rural areas in a region)

ahi = Number of sample enumeration areas allocated to the hth stratum (urban/rural

in the region)

e.g. a11 = 14 for urban area in the Western region

and a12 = 18 for a rural area in the Western region

b = 15 (number of selected households per enumeration area in each stratum)

Then, the first- and second-stage probabilities of selection are:

hi

hihihi

M

MaP

1 and

*2

hi

hiM

bP

Where,

P1 hi is the probability of selecting the ith enumeration area in the hth stratum, and P2 hi is

the probability of selecting a household in the ith enumeration area of the hth stratum. The

overall probability of selection of a household in the ith selected enumeration area of the

hth stratum is the design weight given by:

hihihi PPF 21 *

*

*hi

hi

hi

hihi

M

M

M

baF

The weighting factor (or expansion factor), Whi, for a household in the ith selected

enumeration area in the hth stratum is the reciprocal (inverse) of the overall probability of

selecting that household. That is,

hi

hiF

W1

268

hi

hi

hi

hi

M

M

ba

M **

The number of households successfully interviewed in each enumeration area was used in

the computation. Therefore, the final weight for the sample households in the jth cluster

and in the ith sample PSU in stratum h is given by:

hihi Wb

bW *

''

''

Where:

'b the number of interviews plus the number of no interviews in the sample cluster

''b the total number of interviewed sample households selected in the jth sample PSU

within the ith sample stratum h.

269

Appendix E: Regression tables

Table E1: Determinants of women having experienced domestic violence in the 12

months prior to survey: odds ratios estimated using logistic regressions

Psychological Physical Sexual Social Economic

violence violence violence violence violence

Age category (r:

15–19)

20–24 0.54** 0.50** 0.44* 0.52*** 0.77

(0.13) (0.14) (0.22) (0.12) (0.22)

25–29 0.42*** 0.34*** 0.37** 0.44*** 1.12

(0.11) (0.095) (0.19) (0.10) (0.31)

30–39 0.35*** 0.31*** 0.17*** 0.23*** 0.69

(0.087) (0.095) (0.088) (0.058) (0.19)

40–49 0.21*** 0.12*** 0.033*** 0.15*** 0.40***

(0.068) (0.045) (0.028) (0.045) (0.13)

50–60 0.16*** 0.0086*** 0.020*** 0.11*** 0.27***

(0.054) (0.0090) (0.021) (0.036) (0.088)

Work situation (r:

self-employed)

Employed 0.80 0.76 0.95 1.35 0.73

(0.20) (0.25) (0.39) (0.27) (0.18)

Not working 1.20 1.10 0.81 1.23 1.43**

(0.23) (0.25) (0.25) (0.19) (0.22)

Asset Index 0.95** 0.96 1.01 0.98 0.95**

(0.022) (0.025) (0.011) (0.011) (0.020)

Marital status (r:

never been in a

relationship)

In a monogamous

relationship

0.75

(0.20)

1.55

(0.52)

1863037.8***

(728997.8)

0.89

(0.22)

1.81*

(0.55)

In a polygamous

relationship

1.91**

(0.54)

3.53***

(1.34)

3613973.4***

(1855414.2)

1.47

(0.44)

4.27***

(1.47)

Divorced,

separated or

widowed

1.17

(0.34)

2.14**

(0.80)

1844043.1***

(805891.7)

1.06

(0.30)

2.02**

(0.67)

270

Psychological

violence

Physical

violence

Sexual

violence

Social violence Economic

violence

Education (r: none)

Primary school 1.12 1.82** 1.23 1.82** 1.50**

(0.26) (0.51) (0.71) (0.43) (0.29)

Middle school 1.07 1.35 1.70 1.89*** 1.38

(0.23) (0.37) (0.82) (0.39) (0.28)

Secondary school 1.45 1.44 2.55* 1.71** 1.21

(0.38) (0.52) (1.38) (0.41) (0.32)

Technical school 1.67 1.21 0.75 1.45 1.24

(0.75) (0.79) (0.82) (0.77) (0.55)

Higher education 1.40 1.17 0.93 1.12 0.98

(0.59) (0.63) (0.66) (0.39) (0.39)

Rural area 1.31 0.96 1.00 0.97 1.39**

(0.22) (0.18) (0.30) (0.13) (0.22)

Patriarchal Gender

Norms Index

1.07

(0.081)

0.98

(0.092)

0.89

(0.11)

0.99

(0.068)

0.93

(0.059)

Tolerance to Wife-

beating Index

1.01

(0.069)

1.13

(0.088)

1.01

(0.15)

1.14**

(0.072)

1.09

(0.067)

Acceptance of

Women’s Sexual

Autonomy Index

1.16**

(0.078)

1.17*

(0.10)

1.13

(0.14)

1.16***

(0.066)

1.11*

(0.071)

Decision-making

Power Index

within the

household

0.97

(0.064)

0.93

(0.079)

1.14

(0.14)

0.91

(0.056)

0.95

(0.056)

Respondent

consumes alcohol

1.24

(0.26)

1.35

(0.31)

1.22

(0.44)

0.97

(0.19)

1.39*

(0.26)

Exposure to

violence in the

community

1.77***

(0.24)

1.49**

(0.26)

3.64***

(1.00)

1.88***

(0.25)

1.57***

(0.22)

Witnessed

psychological

violence at home

as a child

2.21***

(0.46)

271

Psychological

violence

Physical

violence

Sexual

violence

Social violence Economic

violence

Witnessed

physical violence

at home as a child

1.73*

(0.49)

Witnessed sexual

violence at home

as a child

1.18

(0.66)

Witnessed social

violence at home

as a child

1.97***

(0.36)

Witnessed

economic violence

at home as a child

2.93***

(0.60)

Regions (r:

Western)

Central 1.63* 1.32 4.77*** 0.97 0.61**

(0.47) (0.60) (2.27) (0.26) (0.15)

Greater Accra 1.50 1.11 3.08** 1.28 0.51***

(0.41) (0.44) (1.70) (0.32) (0.13)

Volta 0.96 0.96 2.00 0.99 0.22***

(0.25) (0.39) (1.13) (0.26) (0.063)

Eastern 2.95*** 2.03* 1.72 1.66* 0.89

(0.83) (0.76) (0.92) (0.44) (0.20)

Ashanti 1.66* 1.68 1.56 1.20 0.70*

(0.49) (0.65) (0.94) (0.29) (0.15)

Brong Ahafo 1.21 1.19 0.55 0.70 0.70

(0.34) (0.48) (0.37) (0.21) (0.16)

Northern 0.70 0.68 0.65 1.13 0.26***

(0.25) (0.31) (0.53) (0.33) (0.079)

Upper East 0.11** 0.14* 0.97 1.10 0.22***

(0.11) (0.15) (1.09) (0.41) (0.087)

Upper West 0.71 1.07 1.64 1.07 0.47*

(0.41) (0.51) (1.76) (0.35) (0.18)

Observations 2,979 2,879 2,979 2,979 2,979

F-test for regional

effects (p-value)

0.14 0.67 0.31 0.75 0.54

Odds ratio; Standard errors in parentheses are clustered within enumeration areas (EA). * p < 0.1; **

p < 0.05; *** p < 0.01


272

Table E2: Determinants of men having experienced domestic violence in the 12 months

prior to the survey: odds ratios estimated using logistic regressions

Psychological

violence

Physical

violence

Sexual

violence

Social

violence

Economic

Violence

Age categories (r:

15–19)

20–24 0.73 0.37** 1.73 0.61 0.86

(0.24) (0.16) (1.69) (0.20) (0.24)

25–29 0.64 0.26** 1.49 0.44** 0.70

(0.25) (0.14) (1.55) (0.18) (0.26)

30–39 0.37** 0.051*** 1.88 0.24*** 0.24***

(0.16) (0.046) (1.87) (0.10) (0.11)

40–49 0.48* 0.059*** 0.52 0.20*** 0.52

(0.20) (0.050) (0.63) (0.088) (0.23)

50–60 0.30** 0.12*** 0.17*** 0.27**

(0.16) (0.092) (0.086) (0.15)

Work situation (r:

self-employed)

Employed 0.91 0.97 2.69* 1.01 2.03***

(0.22) (0.46) (1.59) (0.24) (0.52)

Not working 1.32 0.97 1.68 0.79 2.25**

(0.40) (0.41) (1.22) (0.23) (0.71)

Asset Index 0.99 0.94 1.03* 1.02** 1.00

(0.0089) (0.055) (0.016) (0.0089) (0.019)

Marital status (r:

never been in a

relationship)

In a monogamous

relationship

1.09

(0.29)

0.75

(0.32)

6.02*

(6.18)

2.97***

(0.85)

1.03

(0.30)

In a polygamous

relationship

1.55

(0.66)

0.91

(0.67)

17.4**

(19.9)

5.23***

(2.17)

0.52

(0.33)

Divorced,

separated or

widowed

1.19

(0.37)

0.41

(0.27)

7.91*

(8.51)

1.77*

(0.61)

0.99

(0.34)

273

Psychological

violence

Physical

violence

Sexual

violence

Social

violence

Economic

violence

Education (r: none.

For sexual

violence, r:

Higher)

Primary school 1.87* 1.97 2064392.7*** 1.94 2.87*

(0.68) (1.66) (4702352.7) (0.97) (1.62)

Middle school 1.32 2.18 9448049.2*** 1.73 2.96**

(0.46) (1.63) (26877735.0) (0.74) (1.43)

Secondary 1.91* 4.00* 7752342.1*** 1.54 2.97**

(0.71) (3.12) (16778822.1) (0.71) (1.46)

Technical 1.20 3899351.4*** 2.16 3.70*

(0.73) (9933915.0) (1.35) (2.48)

Higher 1.64 1.81 6917039.8*** 2.23 2.56*

(0.69) (1.91) (3946834.7) (1.09) (1.35)

Rural area 1.28 2.18** 1.18 1.27 1.32

(0.27) (0.75) (0.57) (0.30) (0.31)

Patriarchal Gender

Norms Index

0.92

(0.088)

1.06

(0.22)

0.85

(0.18)

0.84*

(0.083)

1.01

(0.098)

Tolerance to Wife-

beating Index

1.21**

(0.12)

1.47***

(0.19)

1.18

(0.28)

1.04

(0.11)

1.30**

(0.14)

Acceptance of

Women’s Sexual

Autonomy Index

0.91

(0.082)

0.99

(0.16)

1.27

(0.34)

0.96

(0.095)

1.09

(0.11)

Decision-making

Power Index

within household

1.06

(0.13)

1.08

(0.22)

0.85

(0.25)

0.65***

(0.071)

0.88

(0.10)

Respondent

consumes alcohol

1.66***

(0.32)

1.86

(0.72)

2.04*

(0.77)

1.28

(0.26)

1.19

(0.30)

Exposure to

violence in the

community

1.96***

(0.38)

2.26**

(0.77)

2.83***

(1.07)

3.40***

(0.72)

2.62***

(0.51)

274

Psychological

violence

Physical

violence

Sexual

violence

Social

violence

Economic

Violence

Witnessed

psychological

violence at home

as a child

1.80**

(0.51)

Witnessed

physical violence

at home as a child

1.28

(0.58)

Witnessed sexual

violence at home

as a child

2.00

(1.52)

Witnessed social

violence at home

as a child

2.29***

(0.65)

Witnessed

economic violence

at home as a child

2.28***

(0.70)

Regions (r:

Western)

Central 1.44 2.17 1.21 0.79 1.68

(0.53) (1.60) (0.82) (0.28) (0.64)

Greater Accra 1.25 2.39 0.72 1.00 0.80

(0.46) (1.92) (0.49) (0.34) (0.27)

Volta 0.93 0.79 0.60 0.35** 0.20***

(0.32) (0.59) (0.42) (0.15) (0.11)

Eastern 2.43** 2.65 0.44 0.89 1.25

(0.86) (1.96) (0.36) (0.33) (0.47)

Ashanti 1.13 1.51 0.25 0.86 0.89

(0.43) (1.13) (0.21) (0.30) (0.32)

Brong Ahafo 0.35** 1.00 0.27*** 0.86

(0.18) (0.88) (0.12) (0.37)

Northern 0.87 0.24 0.95 0.33** 0.73

(0.35) (0.29) (0.79) (0.15) (0.36)

Upper East 0.35* 1.41 0.16** 0.27*

(0.22) (1.22) (0.12) (0.20)

Upper West 1.48 0.46 0.25

(1.54) (0.26) (0.28)

Observations 1,923 1,904 1,425 1,974 1,974

F-test for regional

effects (p-value)

0.33 0.70 0.78 0.35 0.09

Odds ratio; Standard errors in parentheses are clustered within enumeration areas (EA). * p < 0.1; **

p < 0.05; *** p < 0.01


275

Questionnaire annex

276

Q1: Household questionnaire

Administration Form

IDENTIFICATION

ADDRESS _______________________________________________________

LOCALITY NAME ________________________________________________

NAME OF FIRST PRINCIPAL RESPONDENT ________________________

CLUSTER NUMBER_______________________________________________

STRUCTURE NUMBER____________________________________________

HOUSEHOLD NUMBER___________________________________________

REGION_________________________________________________________

DISTRICT________________________________________________________

URBAN =1 | RURAL = 2

_________________________________________________________________

PERSON INTERVIEWED: Man = 1 | Woman = 2

_________________________________________________________________

INTERVIEWER VISITS

1 2 3 FINAL VISIT

DATE

___________

____________

____________

DAY

MONTH

YEAR

INT.NUMBER

RESULT

INTERVIEWER’S

NAME

SUPERVISOR’S

NAME

RESULT

___________

___________

___________

____________

____________

___________

____________

____________

___________

NEXT VISIT:

DATE

TIME

___________

___________

____________

____________

TOTAL

NUMBER

OF VISITS

RESULT CODES:

1 COMPLETED

2 NO HOUSEHOLD MEMBER AT HOME OR NO ELIGIBLE

RESPONDENT AT HOME AT TIME OF VISIT

3 ENTIRE HOUSEHOLD ABSENT FOR EXTENDED

PERIOD OF TIME

4 POSTPONED

5 REFUSED

6 DWELLING VACANT OR ADDRESS NOT A DWELLING

7 DWELLING DESTROYED

8 DWELLING NOT FOUND

9 OTHER ___________________________

TOTAL PERSONS

IN HOUSEHOLD

TOTAL ELIGIBLE

WOMEN

TOTAL ELIGIBLE

MEN

LANGUAGE OF

RESPONDENT

LANGUAGE OF

INTERVIEW

277

LANGUAGE CODES:

ENGLISH = 1 | AKAN = 2 | GA = 3 | EWE = 4 | NZEMA = 5 | DAGBANI = 6 |

OTHER = 7

Informed consent form

Hello, I am [name], and I am working for Ghana Statistical Service. We are conducting a

national survey about family life and women’s/men’s health.

Your household has been randomly chosen to participate in this survey; and we would

like to ask you and another household member some questions. I want to assure you that

all of your answers will be kept strictly secret. I will not keep a record of your name or

address.

I am using a tablet computer to make the interview easier and quicker, but I will not

record anything or take any pictures.

You have the right to stop the interview at any time, or to skip any questions that you don’t

want to answer. There are no right or wrong answers. We would very much appreciate

your household’s participation in this survey.

The interview takes approximately 45 minutes to complete.

Do you have any questions for me?

May we have your permission to ask you questions,

and are you willing to participate?

UNPROMPTED

YES………………………….1

YES, BUT NOT RIGHT

NOW………………………..2

NO…………………………..0

If no, May I ask why you are not willing to participate?

DO NOT ASK ANY FURTHER QUESTIONS TO

THE RESPONDENT AND THANK HIM/HER FOR

HIS/HER TIME.

1. I don’t have the time

2. My spouse/partner/guardian

does not let me

3. I do not want to talk about my

health and life experiences

4. I do not want to talk to you

5. Other

If 2, when would be more suitable for you?



HIS/HER TIME.

_________________ Date

_________________ Time

278

TO BE COMPLETED BY INTERVIEWER

I certify that I have read the above consent procedure to the participant.

SIGNED: ___________________________________________________

279

1. Household Roster

READ: First of all, I would like to make a comprehensive list of all members of this household. I would therefore like to ask you to give me the names of all

individuals who normally live and eat their meals together here. Please do not include people who have a household elsewhere and come here only for some hours

during the day or to visit temporarily, e.g. servants and visiting family members or friends. Let us start with the head of the household. MAKE SURE THAT

RESPONDENT IS THE HEAD OF THE HOUSEHOLD!

Special Cases: Domestic Servants if they sleep 5 nights a week or more in the household.

Visitors if they have slept in the household for the past 4 weeks.

Please circle ELIGIBLE individual respondents: ANY WOMAN/MAN 15–60 YEARS OF AGE, LIVING IN THE HOUSEHOLD

Q1: Q2: Q3: Q4: Q5: Q6: Q7: Q8:

Person

ID

Name How old is

[name]?

AGE IN

COMPLETED

YEARS

Circle those who

are eligible for the

individual-level

survey

What is

[NAME]’s

sex?

MALE = 1

FEMALE

= 2

What is

[name]’s

relationship

to the head

of the

household?

SEE CODES

BELOW

What is

[NAME]’s

ethnicity?

98=Don’t

know

What is

[NAME]’s

religion?

Is [NAME]

currently

away from

this

household?

Yes = 1

No = 0

THIS COULD

BE FOR

WORK,

EDUCATION

OR

SPENDING

TIME WITH

FAMILY

Can

[NAME]

read or

write in

any

language?

Yes = 1

No = 0

ASK IF

AGE IS 11

OR

OLDER

1

2

3

4

5

6

7

280

8

9

10

11

12

13

14

15

16

17

What is the ID of the respondent? ______________

CODES: Relationships to head Q4

1 = Head

2 = Spouse

3 = Partner

9 = Brother or sister

10 = Niece/nephew by blood

11 = Niece/nephew by marriage

4 = Son or daughter 12 = Other relative

5 = Son-in-law or daughter-in-law 13 = Adopted/foster/step child

6 = Grandchild 14 = Not related

7 = Parent 98 = Don’t know

8 = Parent-in-law

CODES: Religion Q6

1 = No religion

2 = Catholic

3 =

Anglican/Methodist/Presbyterian

4 = Pentecostal/Charismatic

5 = Other Christian

6 = Islam

7 = Traditional/Spiritualist

8 = Buddhist

9 = Other

CODES: Ethnic group Q5

1 = Akan

2 = Ga/Dangme

3 = Ewe

4 = Guan

5 = Mole-Dagbani

6 = Grussi

7 = Gruma

8 = Mande

9 = Other

CODES: Education level Q10

1 = Pre-school

2 = Primary

3 = Middle/JSS/JHS

4 = Secondary/SSS/SHS

5 = TECH/VOC

6 = Higher

98 = Don’t know

CODES: Work situation Q12

1 = Self-employed

2 = Employee

3 = Casual work

4 = Not working and not looking

5 = Not working and looking for work

6 = Student/pupil

7 = Apprentice

8 = Retired/Too old to work

9 = Disabled/sick

10 = Too young to work

98 = Don’t know

281

CODES: Occupation Q13

1 = Managers

2 = Professionals

3 = Technicians and associate

professionals

4 = Clerical support workers

5 = Service and sales workers

6 = Skilled agricultural, forestry and

fishery workers

7 = Craft and related trade

workers

8 = Plant and machine

operators and assemblers

9 = Elementary occupations

10 = Other occupations

CODES: Marital Status Q14

1 = Never married

2 = Living together but not married

3 = Married

4 = Divorced

5 = Separated

6 = Widowed

282

Q9: Q10: Q11: Q12: Q13: Q14: Q15:

Person

ID

Has [NAME] ever

attended school

or is [NAME]

attending school

now?

1= Never Q11

2 = Now

3 = Past

ASK IF AGE IS 3

OR OLDER

What is the

highest level

of school

[NAME] has

attained?

SEE CODES

ABOVE

How many

years of

schooling has

[NAME]

completed at

this level?

ASK IF AGE

IS 3 OR

OLDER

What describes

[NAME]’s

current main

work situation

best?

SEE CODES

ABOVE

ASK IF AGE IS 5

OR OLDER

FOR

INTERVIEWER:

What describes

[NAME]’s main

current occupation

best?

SEE CODES

ABOVE

ASK IF AGE IS 5

OR OLDER

DO NOT READ

OUT LOUT THE

CODES

What is

[NAME]’s

marital

status?

SEE CODES

ABOVE

ASK IF AGE

IS 12 OR

OLDER

How many years has

[NAME] lived in this

locality?

FULL NUMBER OF

YEARS

HAS ALWAYS LIVED

HERE = 991

LESS THAN ONE YEAR

= 992

998 = Don’t know

1

2

3

4

5

6

7

8

9

10

11

12

13

14

283

15

16

17

18

LIST THE IDs OF ALL THE PERSONS IN THE HOUSEHOLD WHO ARE AT LEAST 13 YEARS OLD

2. Employment

Q1: Q2: Q3: Q4: Q5: Q6: Q7: Q8:

Person

ID

Has [NAME]

done any

work over

the last 12

months?

1 = Yes

0 = No

Q13

Over the last

12 months,

how many

months was

[NAME]

involved in

any economic

activities?

EXCLUDING

UNPAID

HOUSEWORK

OR CHORES

Did [NAME]

work during the

last 7 days?

1 = Yes

0 = No

EXCLUDING

UNPAID

HOUSEWORK

OR CHORES

IF NO >>>

NEXT PERSON

IF LAST

PERSON AND

NO>>Q13

Please describe

what work

[NAME] has done

over the last 7

days.

For interviewer:

What work

activities was

[NAME] involved

in during the last

7 days?

SEE

OCCUPATION

CODES

DO NOT READ

OUT LOUT THE

CODES

On

average,

how many

hours per

day was

[NAME]

engaged

in all work

activities

during the

last 7

days?

Was [name]

paid in

cash, in-

kind (self-

consumptio

n) or as a

contributin

g family

member for

any of these

activities?

1 = Yes

0 = No

If No >>

Q11

How was

[NAME] paid

for all the

work

activities

during the

last 7 days?

(multiple

response)

1 = In kind

2 = Cash

3 =

Contributing

family

member

IF 2 >> Q10

If 3 >> Q11

Main

activity

Second

ary

activity

Main

activity

Second

ary

activity

284

1

2

3

4

5

6

7

8

9

10

11

12

Q9: Q10: Q11: Q12: Q13: Q14:

Person

ID

What did

[NAME] receive

in-kind for all

work activities

during the last

month?

1 = Food

2 =

Accommodation

or lodging

3 = Education

(schooling or

vocational)

4 = Clothes

5 = Medicine

6 = Other

How much did [NAME]

earn for all work

activities?

Per:

1 = Day

2 = Week

3 = Month

IF PAID IN-KIND, WHAT

IS THE VALUE?

Who was [NAME]

working for during the

last 7 days? Prompted

1 = Employed by family

member

2 = Private employer

3 = Public sector

4 = Self-employed

within the family

5 = Self-employed

outside the family

6 = Friend

7 = Other

2 ANSWERS

ALLOWED;

Did [NAME] have

a work contract

for the main

activity?

1 = Formal

contract

2 = Verbal contract

0 = No contract

Was [NAME]

involved in house

work/chores (e.g.

cleaning, laundry,

taking care of

children, cooking)

during the last 7

days?

1 = Yes

0 = No

IF NO>>>NEXT

INDIVIDUAL

On average, how

many hours per

day did [NAME]

work on house

work/chores

during the last 7

days?

THEN MOVE TO

NEXT

INDIVIDUAL

285

Main

occupation

Secondary

occupation

Main

occupation

Secondary

occupation

Cash Kind Cash Kind

1

2

3

4

5

6

Q16: Over the last 12 months, has anyone in your household lost their job? 1 = Yes

0 = No >>>then next module

Q17: If yes, then who?

Use household IDs

3. Food consumption and household shocks

Thank you for this information. Now, I would like to know a little bit more about what your household consumes for food.

Name of the item a) Did your household consume

[ITEM] in the last 7 days?

Yes…1

No….0 >>>> Next item

b) On how many days over

the last 7 days did your

household consume this

[ITEM]?

Q1: Cereals

Q2: Beans/pulses/lentils

Q3: Edible oil

Q4: Leafy vegetables

Q5: Any other vegetables

Q6: Tubers (yam, cocoyam, cassava etc.)

Q7: Plantain

Q8: Dairy product, milk

Q9: Eggs

Q10: Meat

286

Name of the item a) Did your household consume

[ITEM] in the last 7 days?

Yes…1

No….0 >>>> Next item

b) On how many days over

the last 7 days did your

household consume this

[ITEM]?

Q11: Fruits

Q12: Fish

Q13: Spices

Q14: Other food (e.g. sugar, honey, tea leaves, coffee)

Q15: During the last lean season, how often have you or a member of your household

gone to bed while still hungry?

1 = more than once a week

2 = once a week

3 = a few times per month

4 = once a month

5 = rarely

6 = never

Q16: During the last harvest season, how often have you or a member of your

household gone to bed while still hungry?

1 = more than once a week

2 = once a week

3 = a few times per month

4 = once a month

5 = rarely

6 = never

Now I would like to ask you some questions related to income shocks that your household might have experienced in the last 12 months.

Over the last 12 months, have you or anyone in your household been affected by any

of the following things?

1 = Yes 0 = No

Q17: Flooding/too much rain

Q18: Drought

Q19: Landslides

Q20: Pest or diseases (crops or livestock)

Q21: Fire (burning down property)

Q22: Insufficient water supply for farming or gardening

287

Q23: How would you or other members of the

household get 1000 Cedis if you really needed to.

For example, because of a medical emergency,

MULTIPLE ANSWERS ALLOWED

1 = Taking a formal loan (from e.g. bank)

2 = Taking an informal loan (from e.g. moneylenders, farmers)

3 = Asking relatives/friends in the same village/town

4 = Asking relatives/friends in different location

5 = Use savings

6 = Use microfinance

7 = Sell durable goods/equipment

8 = Sell land/house

9 = Sell livestock

10 = Take extra work

11 = Other

288

4. Household Assets and Decision Making

Now I would like to know about goods that belong to you or any other household member. Please also consider goods that are part of your

business.

How many of the following things does your household possess

and who (mainly) owns and uses the respective item(s)?

A. OWN

Yes = 1, No = 0

B. Owner(s): Use hhd IDs

MULTIPLE IDs ALLOWED

Q1: Bicycle

Q2: Motorcycle or motor scooter

Q3: Tricycle

Q4: Animal-drawn cart

Q5: Car

Q6: Truck

Q7: Boat with a motor

Q8: Boat without a motor

Q9: Tractor

Q10: A) How much agricultural land does your household possess?

If 0, skip to Q13

B) Unit:

1= Hectares

2 = Acres

3 = Poles

4 = Plots

C) Who owns the agricultural land? Household IDs

D) Who usually make decisions regarding the use of agricultural land

THIS INCLUDES DECISIONS ON WHAT TO GROW AND SALE AND RENT

Household IDs

Q11: Is part of this land cultivated by a tenant? Yes=1, No=0 >>> Q13

Q12: For the plots which you did not crop yourself, what net rent did

you receive from the tenant during the last year? IF RENT IS

PAID IN KIND, ASK THE RESPONDENT TO ESTIMATE THE

MONETARY VALUE

Total amount in Cedis

____________________

98 = Don’t know

Q13: How much agricultural land does your

household cultivate?

1= Hectares

2 = Acres

3 = Poles

289

INCLUDE CULTIVATED LAND THAT IS

RENTED IN AND EXCLUDE LAND THAT IS

RENTED OUT. If 0 >> NEXT MODULE

4 = Plots

Q14: How much of this cultivated land is irrigated?

If 0 >>Q16

1= Hectares

2 = Acres

3 = Poles

4 = Plots

Q15: What is the principal source of

water for this agricultural land?

1 = Dam

2 = Hole with a pump

3 = Well

4 = Natural spring

5 = River/lagoon/lake

6 = Storage tank

7 = Rainfall

8 = Other

Q16: IF HOUSEHOLD OWNS AND CULTIVATE LAND, What were

your net earnings of the last 12 months from the land you own

and cultivate? IF HARVEST IS NOT SOLD TO THE MARKET,

ASK THE RESPONDENT TO ESTIMATE THE MONETARY

VALUE


______________________

98 = Don’t know

Q17: Does your household cultivate land owned by someone outside

your household?

1=Yes

0=No >>> NEXT MODULE

Q18: What was the net earnings of the last 12 months for this rented

land? IF HARVEST IS NOT SOLD TO THE MARKET, ASK THE

RESPONDENT TO ESTIMATE THE MONETARY VALUE


________________________

98 = Don’t know

For agricultural purposes: IF HOUSEHOLD DOES NOT OWN OR CULTIVATE LAND, SKIP TO Q32

How many of the following things does your household possess and who

(mainly) owns the respective item(s)?

A. How many? B. Owner(s): Use hhd IDs

MULTIPLE IDs ALLOWED

Q19: Plough

Q20: Cart

Q21: Thresher

Q22: Trolley

Q23: Water pump

Q24: Irrigation pump

Q25: Grain storage bin

Q26: Sprayers

290

Q27: Cattle

Q28: Milk cow and bulls

Q29: Horses, donkeys, mules

Q30: Goats

Q31: Sheep

Q32: Pigs

Q33: Rabbits

Q34: Grasscutter

Q35: Poultry

Q36: Who in your household makes decisions regarding the use of the following animals:

a) Cattle

b) Milk cows and bulls

c) Horses, donkeys and mules

d) Goats

e) Sheep

f) Pigs

g) Rabbits

h) Grasscutter

i) Poultry

Q37: Does anyone in your household have a bank account? Who? Yes = 1, No = 0

Q38: Does anyone in your household use a microfinance scheme? Who? Yes = 1, No = 0

Q39: Do you or any member of your household receive remittances from people

living outside your household?

Yes = 1

No = 0 >>>Q41

Q40: What is the value of these remittances? ____________________

1 = Day 3 = Month

2 = Week 4 = Year

Q41: Are you or any member of your household a participant in or beneficiary of a

cash transfer programme by the government? (LIST) THIS DOES NOT

INCLUDE CASH TRANSFERS FROM FRIENDS OR FAMILY

Yes = 1

No = 0 >>Q43

Q42: What is the value of the transfer? _______________________

291

THIS DOES NOT INCLUDE CASH TRANSFERS FROM FRIENDS OR FAMILY 1 = Day 3 = Month

2 = Week 4 = Year

I would like to ask you about spending decisions in your household.

How are spending decisions made in your household? Who usually… Decision-maker (Use Household IDs)

0 = None of us

199 = All of us

98 = Don’t know

99 = Not applicable

Q43: Decides how the earnings of the head of the household are used?

Q44: Decides how the money from the remittances is used?

Q45: Decides how the money from the cash transfer programme is used?

Who decides how to spend money on the following items?

Q46: Food

Q47: Education for the children

Q48: Healthcare for the children

Q49: Healthcare for yourself

Q50: Clothing

Q51: Purchase for daily household needs

Q52: Major household purchases

Q53: Do you own or rent the dwelling that you

live in?

1= Owner occupied

2 = Renting >>>>>Q55

3 = Rent free >>>>>Q55

4 = Perching >>>>>Q55

5 = Squatting >>>>>Q55

Q54: Who owns it? Roster IDs

Q55: H

0

1

I would like to ask you a few questions

about your housing condition.

What is the main source of drinking water

for members of your household?

1 = Piped into dwelling/indoor

2 = Piped to yard/plot

3 = Public tap/standpipe

4 = Tube well or borehole

5 = Protected well

6 = Unprotected well

7 = Protected spring

9 = Rainwater

10 = Tanker truck

11 = Cart with small tank

12 = Surface water (river/dam/lake/pond

/stream/canal/irrigation channel)

13 = Bottled water

14 = Sachet water

292

8 = Unprotected spring 15 = Other

Q56: Which of the following does your

household use the most:

1 = A private toilet

2 = A shared toilet with other families

3 = A public toilet

4 = None of these

Q57: H

0

2

What main kind of toilet

facility do members of your

household usually use?

1= Flush to piped sewer system

2 = Flush to septic tank

3 = Flush to pit latrine

4 = Flush to somewhere else

5 = Flush, don’t know where

6 = Ventilated improved pit latrine

7 = Pit latrine with slab

8 = Pit latrine without slab/open pit

9 = Bucket/pan

10 = Composting toilet

11 = No facility/bush/field/sea

Q58: H

0

3

Which material is mainly used

for the roof of your house?

1 = No roof

2 = Thatch/palm leaf/sod

3 = Palm/bamboo

4 = Wood planks

5 = Cardboard

6 = Metal

7 = Wood

8 = Calamine/cement fibre

9 = Ceramic tiles/brick tiles

10 = Burnt bricks

11 = Cement

12 = Roofing shingles

13 = Asbestos/slate

14 = Other

Q59: Which material is mainly used

for the exterior walls of your

house?

1 = No walls

2 = Cane/palm/trunks

3 = Dirt/mud

4 = Bamboo with mud

5 = Stone with mud

6 = Uncovered adobe

7 = Plywood

8 = Cardboard

9 = Reused wood

10 = Cement or cement blocks

11 = Stone with lime/cement

12 = Bricks

13 = Covered adobe

14 = Wood planks/shingles

Q60: Which material is mainly used

for the floor of your house?

1 = Earth/sand

2 = Dung

3 = Wood planks

4 = Palm/bamboo

5 = Parquet or polished wood

6 = Vinyl or asphalt strips

7 = Ceramic tiles/terrazzo

8 = Cement

9 = Woollen carpet/synthetic carpet

10 = Linoleum/rubber carpet

Q61: How many rooms do you have in the dwelling that you live in?

Q62: How many rooms do you use for sleeping in the dwelling that you live in?

293

5. Household Assets and Problems in the Community

Now I would like to ask you about your source(s) of electricity.

On a typical day, how many hours

per day do you use this source of

energy for lighting?

Q24: Grid electricity (household is a utility

consumer)

Q25: Generator (from neighbour or relative)

Q26: Generator (personal)

Q27: Mini-hydro

Q28: Solar home system (e.g. solar panel)

Q29: Storage (car) battery

Q30: Candles

Q31: Kerosene (wick and hurricane lamps)

Q32: LPG

Finally, I would like to ask you some questions about your community.

Q33: On a typical day, in your community, how many hours of electricity

is available?

Do you know if any of the following are problems in your

community or neighbourhood?

Yes = 1 No = 0

Q34: Chieftaincy disputes

Q35: Sales of drugs (illicit or hard drugs)

Q36: Drug abuse (illicit or hard drugs)

Q37: Alcohol abuse

Q38: Domestic Violence

Q39: U Poor/unreliable water provision

Q40: Poor/unreliable electricity provision

Q41: Poor/unreliable waste removal

Q42: Poor roads

Q43: Poor health service provision

Q44: Is there a shortage of water in this community

during the dry season?

1 = Yes

0 = No >>> Q52

2 = There is always dry season >>>Q49

Does your household have: 1 = Yes

0 = No

1 = Yes

0 = No

Q1: Radio Q2: TV

Q3: Mobile phone Q4: Bed

Q5: Other furniture Q6: Cooking utensils

Q7: Washing machine Q8: Fan

Q9: Air conditioner Q10: Refrigerator

Q11: Iron Q12: Sewing machine

Q13: Kerosene stove Q14: Gas/LPG/biogas stove

Q15: Electric stove Q16: Open stove/open fire/coal pot

Q17: Dish/cable TV Q18: Wall clock

Q19: Landline telephone Q20: Digital photo-camera

Q21: Tablet Q22: Computer

Q23: Mosquito nets

294

1 = More

than

once a

week

2 =

Once

a

week

3 = A

few

times a

month

4 =

Once a

month

5 =

Rarely

(once in

a

while)

6 =

Never

Q45: How often is the shortage of

water in this community


Q46: How often do disputes over

water occur between

members of this community



water occur between


and a different community


Q48: How often is the shortage of

water in this community?


water occur between

members of this

community?


water occur between


and a different community?

Q51: Is there a shortage of agricultural land in this

community?

1 = Yes, 0 = No >>> Q56,

99 = N.A. >>> Q56


agricultural land occur

between members of this

community?

Q53: How often do arguments

over grazing land occur in

this community? If not

applicable, because there is

no grazing land, enter 99

Q54: How often do arguments

over grazing land occur

between members of

different communities? If not

applicable, because there is

no grazing land, enter 99

Q55: Are mobile phone networks available

in all areas of this community?

1 = All

2 = Most

3 = Some

4 = Few

5 = None

295

Q56: Which method of transport would you

primarily use to go to the district capital?

CIRCLE THE APPROPRIATE AND

MAIN MEAN OF TRANSPORT

1 = Foot

2 = Car

3 = Bicycle

4 = Motorcycle

5 = Tricycle

6 = Boat/canoe

7 = Public transport (e.g.

bus or tro-tro)

8 = Truck

9 = Animal drawn cart

10 = Tractor

98 = Don’t know

Q57: How long does it take to go to the district capital using this

method of transport in minutes?

1 = Minutes

2 = Hours

Thank you very much for your time. I would like to interview another member of your household

to go into more detail about issues related to health and society.

Could I take your phone number for verification purposes? __________________________

Situation of respondent:

1 Alone

2 Children were present some of the time (6–14 years old)

3 Children were present all of the time (6–14 years old)

4 Adults were present some of the time (above 14 years old)

5 Adults were present all of the time (above 14 years old)

66 Other

296

Selection of Respondent for the Individual Interview

Look at the last digit of the household questionnaire serial number on the cover page. This is the

row number you should go to. Check the total number of eligible females or males in the household

schedule. How many are eligible? ___________

This is the columnyou should go to. The cell where the row and the column meet is the number

of the selected woman or man for the domestic violence module in the household schedule.

For example, the household was selected to interview a woman and there are three eligible women

age 15–60 (line IDs 02, 04 and 05). If the household questionnaire serial number is ‘216’, the last

digit is ‘6’; therefore, go to row ‘6’. There are three eligible women age 15–60 in the household,

therefore go to column ‘3’. Follow the row and column and find the number where the row and

column meet (‘2’) and circle the box. Now go to the household schedule and find the second woman

who is eligible for the woman's interview (line number ‘04’ in our example). Write her line ID and

name (if given) below.

TABLE FOR SELECTION OF RESPONDENT

LAST DIGIT OF

THE HOUSEHOLD

QUESTIONNAIRE

SERIAL NUMBER

TOTAL NUMBER OF ELIGIBLE WOMEN 15–60 /

MEN 15–60 IN THE HOUSEHOLD

1 2 3 4 5 6 7 8

0 1 2 2 4 3 6 5 4

1 1 1 3 1 4 1 6 5

2 1 2 1 2 5 2 7 6

3 1 1 2 3 1 3 1 7

4 1 2 3 4 2 4 2 8

5 1 1 1 1 3 5 3 1

6 1 2 2 2 4 6 4 2

7 1 1 3 3 5 1 5 3

8 1 2 1 4 1 2 6 4

9 1 1 2 1 2 3 7 5

Line ID of selected woman/man: _________

Name of selected woman/man: ___________________________________________________

ASK IF YOU CAN TALK WITH THE SELECTED WOMAN/MAN for the second part of the

questionnaire. IF SHE/HE IS NOT AT HOME, AGREE ON A DATE FOR A RETURN VISIT.

297

Q2: Individual questionnaire Administration Form

IDENTIFICATION

ADDRESS _________________________________________________________

LOCALITY NAME _________________________________________________

NAME OF FIRST PRINCIPAL RESPONDENT _________________________

CLUSTER NUMBER _______________________________________________

STRUCTURE NUMBER ____________________________________________

HOUSEHOLD NUMBER ___________________________________________

REGION _________________________________________________________

DISTRICT ________________________________________________________

URBAN =1 | RURAL = 2

_________________________________________________________________

PERSON INTERVIEWED: Man = 1 | Woman = 2

_________________________________________________________________

INTERVIEWER VISITS

1 2 3 FINAL VISIT

DATE

___________

____________

____________

DAY

MONTH

YEAR

INT. NUMBER

RESULT

INTERVIEWER’S

NAME

SUPERVISOR’S

NAME

RESULT

___________

___________

___________

____________

____________

___________

____________

____________

___________

NEXT VISIT:

DATE

TIME

___________

___________

____________

____________

TOTAL

NUMBER

OF VISITS

RESULT CODES:

1 COMPLETED

2 THE FIRST RESPONDENT DID NOT AGREE TO THE INTERVIEW OF THE SECOND

RESPONDENT

3 REFUSED BY THE SECOND RESPONDENT

4 SECOND RESPONDENT NOT AVAILABLE

5 COULD NOT PROCEED BECAUSE OF LACK OF PRIVACY

6 OTHER ___________________________

LANGUAGE OF

RESPONDENT

LANGUAGE OF

INTERVIEW

298

LANGUAGE CODES:

ENGLISH = 1 | AKAN = 2 | GA = 3 | EWE = 4 | NZEMA = 5 | DAGBANI = 6 |

OTHER = 7

Was the respondent alone during the following domestic violence

modules? 1 = Yes 0 = No

Social violence or controlling behaviour

Sexual violence

Physical violence

Psychological violence

Economic violence

299

Hello, I am [name], and I am working for Ghana Statistical Service. We are conducting a

national survey about women’s/men’s health and life experiences.

Your household has been randomly chosen to participate in this survey; and we would

like to ask you some questions.

I want to assure you that all of your answers will be kept strictly secret. I will not keep a

record of your name or address. You have the right to stop the interview at any time, or to

skip any questions that you don’t want to answer. There are no right or wrong answers.

Some of the topics may be difficult to discuss, but many women/men have found it useful

to have the opportunity to talk.

We would very much appreciate your participation in this survey.

Do you have any questions?

(The interview takes approximately 1 hour and 30 minutes to complete.)

May I have your permission to ask you questions, and

are you willing to participate?

YES………………………….1

NOT NOW………………….2

NO…………………………..0

If no, May I ask why you are not willing to participate?



HIS/HER TIME.

1. I don’t have any time

2. My spouse/partner/guardian

does not let me

3. I do not want to talk about my

health and life experiences

4. I do not want to talk to you

5. Other

If it is not a good time to talk, when would be more

suitable for you?



HIS/HER TIME.

_________________ Date

_________________ Time

Q1: For interviewers: What is sex of

respondent?

1 = Male, 2 = Female

TO BE COMPLETED BY INTERVIEWER

I certify that I have read the above consent procedure to the participant.

SIGNED: _______________________________________________________

300

1. Safety and Social Capital

First, I would be interested to know a bit more about how you feel living in your community.

Q2: How safe do you feel in the area you

live in?

1 = Very safe

2 = Somewhat safe

3 = Somewhat unsafe

4 = Very unsafe

Q3: Why don’t you feel

safe in the area you

live in?

1. Fear of being robbed

2. Fear of

burglary/vandalism while

being away

3. Fear of sexual

harassment/rape

4. Fear of verbal harassment

5. Fear of physical

harassment/attacked

6. Fear of being insulted/ridiculed

7. Fear of something else

Q4: How safe do you

think women and

girls in this area feel

when going out

alone?

During the day During the night

1 = Very

safe

2 =

Somewhat

safe

3 = Somewhat

unsafe

4 = Very

unsafe

1 = Very safe

2 =

Somewhat

safe

3 = Somewhat

unsafe

4 = Very unsafe

Q5: How safe do you

think men and boys in

this area feel when

going out alone?

During the day During the night

1 = Very

safe

2 =

Somewhat

safe

3 = Somewhat

unsafe

4 = Very

unsafe

1 = Very safe

2 =

Somewhat

safe

3 = Somewhat

unsafe

4 = Very unsafe

Q6: Q7: Q8:

Have you

ever

witnessed

an act

related to

[problem]

over the last

12 months?

1 = Yes,

0 = No

Have you ever

personally

been affected

by or

experienced

acts related to

[problem] over

the last 12

months?

1 = Yes

0 = No >>>

Next item

How were you affected?

1 = Damage to

personal property

2 = Loss of

personal property

3 = Physical harm

4 = Psychological

harm, stress or

worry

5 = Loss of

income

6 = Damage

to communal

property

7 = Fear of

leaving the

house

8 = Other

A) Theft, robbery, burglary

B) Vandalism

C) Arson

D) Kidnapping/abduction

E) Extortion

F) Local bullies

G) Political/communal

violence

H) Highway attacks

I ) Sexual harassment

G) Rape

K) Defilement

301

L) Seduction

Q9: If a group of people in your neighbourhood decided to organise

a neighbourhood project, such as a sanitation or clean-up project,

would you like to contribute (time or money) to this project?

1 = Yes, happily

2 = Yes, I will

have to

0 = No

Q10: How many marriage ceremonies (weddings, engagements) have

you been personally invited to in the last 12 months?

________________

Q11: How many funerals have you attended in the last 12 months? ________________

Q12: Are you a member of any of the following groups, organisations and associations?

Prompt

1 = Area residents’ association

2 = Cooperative

3 = Professional association

4 = Trade union

5 = Credit/finance group

6 = Neighbourhood council

(including landlords’ association)

7 = School committee or parent’s

group (PTA)

8 = Sports group or club,

youth group

9 = NGO or civic group

10 = Wives’ association or

club

11 = Fan/social club

12 = Cultural association (e.g.

theatre group, reading

group)

13 = Political group or

association

14 = Women’s group

15 = Religious group

16 = Family committee

17 = Ethnic or tribal

association

18 = Student

association (Old Boy’s

club)

19 = Other

Q13: If a member of your household were to get into a disagreement with someone from

this community, who would they approach to resolve it?

1 = No one, they will solve it

amongst themselves

2 = Their friends

3 = Family members

4 = Police

5 = Other government

officials

6 = Community

leaders

7 = Chief/Queen

mother

8 = The district

assembly

9 = Religious leaders

10 = Other members of the

community

11 = Other

12 = Don’t know

302

2. Safety in the Home and Controlling Behaviour

Q1: How safe do you feel in your

home?

1 = Very safe

2 = Somewhat safe

3 = Somewhat unsafe

4 = Very unsafe

The next questions are about things

that happen to many men and

women. Please can you tell me if

anyone has ever (even when you

were a child) done or tried to do the

following things to you:

Clarify: Anyone includes your partner,

members of your household, family,

work colleagues etc.

a) Kept you

from seeing

your friends

or family of

birth

b) Stopped you

from leaving

your house?

c) Insisted on

knowing where

you are at all

times including

by controlling,

checking your

phones, texts or

emails

d) Stalked you,

including by

calling,

messaging,

watching or

following you to

a point where

you felt

uncomfortable

e) Spread false

information

about you or

distributed

photos or

videos of you

without your

permission

f) Forced you to

have an

abortion

Q2: 1 = Yes

2 = Someone tried >>> NEXT

ITEM

0 = No >>> NEXT ITEM

Q3: Has this happened in the last

12 months?

1 = Yes, 0 = No >>> NEXT

ITEM

Q4: In the last 12 months, was the

person who did this to you

from your current household?

1 = Yes all of them, 2 = Some

of them, 0 = No >>Q6

Q5: In the last 12 months, who did

this to you from your current

household? Use IDs from the

303

household roster

Then skip to Q8


this to you? Use codes below

Q7: What was the sex of this

person?


Q8: Over the last 12 months, how

often has this happened? Use

codes below

Q9: When you were a child, did you ever observe this kind of behaviour between

your parents or other members of your household?

1 = Yes, parents

2 = Yes, other members

0 = No

Q10: Think of 10 of your closest neighbours, how many of them do you know, or

think, experienced this kind of behaviour over the last 12 months?

98 = Don’t know

_______________ out of 10

CODES FOR Q6 and 16: WHO DID THIS TO YOU? AND WHO DID YOU DO THIS TO? CODES FOR Q8 and 13: HOW OFTEN HAS

THIS HAPPENED?

304

1 = Partner

2 = Former partner

3 = In-laws/partner’s family

4 = Parents

5 = Sons, daughters

6 = Brothers, sisters

7 = Uncles, aunts

8 = Grandparents

9 = Cousins, nieces, nephews, other

relatives

10 = Priest or religious leader

11 = Other relatives who do not live in this household

but in the community

12 = Other relatives who live outside this community

13 = Friend

14 = Teacher

15 = Employer

16 = Classmates

17 = Work colleagues

18 = Police

19 = Other people in the community

20 = Other people outside the community

1 = Once only

2 = Once in a while

3 = Less than once per

month

4 = Once per month

5 = 2–3 times per month

6 = Once a week

7 = 2–3 times a week

8 = Almost every day

9 = Every day

Please can you tell me if

you have ever done or tried

to do the following things

to someone else:

Clarify: Anyone includes your

partner, members of your

family, work colleagues etc.

a) Kept

someone from

seeing his/her

friends or

family of birth

b) Stopped

someone from

leaving their

home

c) Insisted on

knowing where

someone is at all

times including by

controlling or

checking their

phones, texts or

emails

d) Stalked

someone

including by

calling, texting,

watching or

following to a

point where you

made them feel

uncomfortable

e) Spread false

information about

someone and/or

distributed

photos or videos

of someone

without their

permission

f) Forced someone

to have an abortion

Q11: 1 = Yes

2 = I tried >>> NEXT

ITEM

0 = No >>> NEXT

ITEM

305

Q12: Has this happened in

the last 12 months?

1 = Yes, 0 = No >>>

NEXT ITEM

Q13: Over the last 12

months, how often has

this happened?

Use codes above


months, was the

person you did this to

from your household?

1 =Yes, 0 = No

>>>>Q16


months, who did you

do this to from your

household? Use

Household IDs


months who did you

do this to?

Use codes above

Q17: What was the sex of

this person?


306

3. Attitudes

I would like to know more about how people sometimes relate to each other and your thoughts

about it. We have prepared 4 short stories and would like to know your reaction. Please keep in

mind that there are no right or wrong answers.

S1: After a marital dispute the wife refused to talk to her husband, and would refuse to make his

dinner or any other household chores. The husband hit his wife, but without leaving a mark.

Q1: How acceptable do you think the

behaviour of the husband is?

1 = Completely

unacceptable

2 = Somewhat

unacceptable

3 = Somewhat

acceptable

4 = Completely

acceptable

98 = Don’t know/Can’t

say

Q2: Do you think it is the woman’s fault? 1 = Yes, mostly

2 = Partially

3 = Not at all

S1’: After a marital dispute the wife refused to talk to her husband, and would refuse to make

his dinner or any other household chores. The husband hit his wife, but without leaving a mark.

The local priest/Imam said that it was the woman’s fault because she didn’t fulfil her duties.

Q1: How acceptable do you think the

behaviour of the husband is?

1 = Completely

unacceptable

2 = Somewhat

unacceptable

3 = Somewhat

acceptable

4 = Completely

acceptable


say

Q2: Do you think it is the woman’s fault? 1 = Yes, mostly

2 = Partially

3 = Not at all

Q3: How likely do you think it is that

your local priest/Imam would react

like this?

1 = Very likely

2 = Somewhat likely

3 = Somewhat

unlikely

4 = Very unlikely


say

S2: A newly married man in your community, who is usually a very calm and generous man,

has been facing some financial problems. One day his boss informs him that he has lost his job.

He comes home after this, has an argument with his wife, and beats her up to the point that she

passes out/becomes unconscious. This is the first time an incident like this happens. The wife

later reports the incident to the police, and the man is arrested.

Q4: Do you agree that the woman did the

right thing by going to the police?

1 = Fully agree

2 = Somewhat agree

3 = Somewhat

disagree

4 = Fully disagree


say

Q5: If 3–4, why do

you think the

woman should

not have gone to

the police?

1 = Because it was the first time it

happened

2 = Because such matters should

remain in the family

3 = Because this is a normal

behaviour

4 = Because it will break the

family

5 = Because it will create

financial problems

6 = Other

307

Q6: Do you agree that the police did the

right thing by arresting the man?

1 = Fully agree

2 = Somewhat agree

3 = Somewhat disagree

4 = Fully disagree


say

Q7: If 3–4, why do you

think the police

should not have

arrested the man?


happened




behaviour


family


financial problems

6 = Because the punishment is

too harsh

7 = Other

Q8: Q

8

:

What do you think

the most adequate

punishment for the

husband would be?

1 = Financial compensation to the

wife or her family

2 = Discussions between the wife

and husband’ s families

3 = Sentencing to jail

4 = Official warning by the

police/justice

5 = Police should be physically

rough with him

6 = Apology

7 = Fine

8 = Husband should be bonded

to a note of good behaviour

9 = It is enough to just arrest

him

10 = No punishment needed

S2’: A woman repeatedly makes demeaning comments to her husband about him not making

enough money, and not being able to provide for a nice home and gifts for her. She constantly

compares him to his brother, who is a successful businessman. One evening, after the woman

made such comments and the couple get into an argument, the man beats her up to the point

that she passes out/becomes unconscious. This is the first time an incident like this has

happened. The wife later reports the incident to the police, and the man is arrested.

Q1: How common do you think it is that a

woman treats her husband like this?

1 = Very common

2 = Somewhat

common

3 = Somewhat

uncommon

4 = Very uncommon

Q2: Do you agree that the woman did the

right thing by going to the police?

1=Fully agree

2= Somewhat agree

3=Somewhat disagree

4=Fully disagree


say

Q3: If 3-4, why do you

think the woman

should not have

gone to the police?


happened




behaviour

4 = Because it will break the family


financial problems

6 = Because the woman is the

one to blame

7 = Other

Q4: Do you agree that the police did the

right thing by arresting the man?

1 = Fully agree

2 = Somewhat agree

3 = Somewhat

disagree

4 = Fully disagree


say

308

Q5: If 3–4, why do you

think the police should

not have arrested the

man?

1 = Because it was the first

time it happened

2 = Because such matters

should remain in the family


behaviour


family


financial problems

6 = Because the punishment is

too harsh

7 = Because the man is not to

blame

8 = Other

Q6: What do you think the

most adequate

punishment for the

husband would be?

1 = Financial compensation to

the wife or her family

2 = Discussions between the

wife and husband’ s families

3 = Sentencing to jail

4 = Official warning by the

police/justice

5 = Police should be

physically rough with him

6 = Apology

7 = Fine

8 = Husband should be bonded

to a note of good behaviour

9 = It is enough to just arrest him

10 = No punishment needed

S3: Suppose you hear a dispute between your closest neighbour and her young son. You then

witness the son leave the house crying and with several injuries on his face. You know that there

are two other young children in the house, and you can still hear your neighbour screaming.

Q7: What do you

think you should

do?

MULTIPLE

ANSWERS

ALLOWED

1 = Nothing, I don’t think it is a

problem

2 = Nothing, it is not my

business

3 = Approach the police and ask

them to deal with it

4 = Approach religious

authorities and ask them to deal

with it

5 = Approach community

leaders and ask them to deal

with it

6 = Go directly to your neighbour’s

house and try to stop the dispute

and calm down your neighbour

7 = Approach a relative of the

neighbour and ask them to deal

with it

8 = Approach other members of the

community for help.

9 = I would go and check what

happened

S4: Suppose a man in your community makes lewd comments to his niece and sometimes

touches her inappropriately. This man is a community leader, and talks about the girl

inappropriately in front of other people, making comments about her revealing clothes and

body. The girl’s mother asks her to wear more modest clothes whenever the uncle visits but

does nothing further.

Q8: What do you

think the girl

should do?

MULTIPLE

ANSWERS

ALLOWED

1 = She should dress more

modestly

2 = She should stand up for

herself/ask the uncle to stop

3 = She should threaten the uncle

that she will tell the police or other

authorities

4 = She should tell her father

5 = She should tell other family

members

6 = She should tell religious leaders

7 = She should tell the police or

other authorities

8 = The girl should take this as a

compliment

9 = This is what men do; she should

avoid being with him

309

Q9: What do you

think the

mother should

do?

1 = The mother is right to ask her

daughter to dress modestly

2 = She should ask the uncle to stop

3 = She should threaten the uncle

that she will tell the police or other

authorities

4 = She should tell other family

members

5 = She should tell religious leaders

6 = She should tell the police or

other authorities

7 = She should advise her daughter

to go to the police/authorities

8 = She should tell the girl to take

this as a compliment

9 = This is what men do; she should

advise/help her daughter to avoid

being with him

S5: Suppose that someone that you know, who has been dating his girlfriend for two years now,

starts to become very paranoid that she is being unfaithful to him. This started when the

girlfriend mentioned that a young man has joined the office where she works. Ever since then,

he checks the messages on her phone, follows her when she is with her friends and shows up at

her work place unannounced. He also demands to know where she is at all times, and they get

into arguments whenever she comes home from work late.

Q10: What do you think of the

man’s reaction?

1 = He has the right to do this

2 = I sympathise with him, but he should not do this

3 = He should never do this

S5’: Suppose that someone that you know, who has been dating her boyfriend for two years

now, starts to become very paranoid that he is being unfaithful to her. This started when the

boyfriend mentioned that a young woman has joined the office where he works. Ever since

then, she checks the messages on his phone, follows him when he is with his friends and shows

up at his work place unannounced. She also demands to know where he is at all times, and they

get into arguments whenever he comes home from work late.

Q11: What do you think of the

woman’s reaction?

1 = She has the right to do this

2 = I sympathise with her, but she should not do this

3 = She should never do this

READ: I am now going to read you a list of statements, and I would like you to tell me whether

you generally agree or disagree with the statement. There are no right or wrong answers.

1 = Fully

agree

2 =

Somewhat

agree

3 =

Somewhat

disagree

4 = Fully

disagree

98 =

Don’t

know/

Can’t

say

Q12: A good wife obeys her

husband even if she

disagrees with him

Q13: Family problems should only

be discussed with people in

the family

Q14: In order to bring up, raise or

educate a child properly, the

child sometimes needs to be

physically punished

310

Q15: I If a woman wears revealing

clothes, it is okay for men to

make some sexual comments

as she walks in her

neighbourhood

Q16: If a woman wears revealing

clothes, it is her fault if she is

raped

Q17: It is allowed for men to stare

at women they don’t know

Q18: In your opinion, does a man have a good reason to hit

his wife if: Yes No Don’t know

a) She does not complete her household work to his

satisfaction

1 0 98

b) She disobeys him 1 0 98

c) She refuses to have sex with him 1 0 98

d) She asks him whether he has other girlfriends 1 0 98

e) He suspects that she is unfaithful 1 0 98

f) He finds out that she has been unfaithful

g) She contracts a sexually transmitted disease

h) She goes out without telling him

i) She neglects the children

j) She argues with him

k) She burns the food

1

1

1

1

1

1

0

0

0

0

0

0

98

98

98

98

98

98

Q19: In your opinion, can a married woman refuse to have

sex with her husband if: Yes No Don’t know

a) She doesn't want to 1 0 98

b) He is drunk 1 0 98

c) She is sick 1 0 98

d) He mistreats her 1 0 98

e) She is menstruating

f) He does not want to use a condom/contraception

g) She finds out that he has girlfriends

1

1

1

0

0

0

98

98

98

4. Relationships and Reproductive Decisions

I want to ask a few questions about you and your current or most recent partner. By partner we

mean someone that you have a relationship with, for example a husband or wife, girlfriend or

boyfriend, lover or other romantic relationship.

Q1: Have you ever lived with a partner? 1 = Yes

0 = No >>>> Q3

Q2: How many partners have you ever lived with?

INCLUDE PARTNERS THE RESPONDENT IS

MARRIED TO AND THOSE YOU ARE

CURRENTLY LIVE WITH

_________________ times

Q3: Have you ever been married?

1 = Yes

0 = No >>>> Q5

311

THIS INCLUDES CURRENT MARRIAGE

Q4: How many times have you ever been married?

_________________ times

Q5: Are you currently in a relationship? 1 = Yes

0 = No >>>>> skip to Q52

Q6: Are you in a relationship with more than one

person?

1 = Yes

0 = No >>>>> skip to Q10

Q7: If yes, how many persons are you in a relationship

with?

THIS INCLUDES ALL RELATIONSHIPS, EVEN

THE PRIMARY RELATIONSHIP

__________________ number

Q8: Are you married to any of these other persons? 1 = Yes

0 = No

Q9: If yes, how many of these persons are you married

to?

THIS INCLUDES ALL MARRIAGES, EVEN THE

PRIMARY MARRIAGE

For the remainder of these questions I would like to ask you about your primary/main

partner. This could be the person that you live with.

Q10: Are you in a relationship with a man or a woman? 1 = Man

2 = Woman

Q11: How old were you when you entered this

relationship?

_________________ years

Q12: Was this an approximation 1 = Yes

0 = No

Q13: How did you get to know your

current partner?

1 = Arranged meeting

2 = We are family (e.g. cousin)

3 =

Friends/classmates/colleagues/neighbours/customer

4 = At party, festival or other social event

5 = Other

Q14: Who influenced the decision of

choosing your partner?

MULTIPLE ANSWERS

ALLOWED

1 = Myself

2 = Parents

3 = Extended family

4 = Church/Mosque/cult/sect

5 = Other

If respondent is man then skip to Q17

Q15: For women: Does your partner have any wives in

addition to your relationship?

THIS INCLUDES THE RESPONDENT IF S/HE IS

MARRIED

1 = Yes

0 = No >>>>> Q17

98 = Don’t know >>>>> Q17

Q16: For women: How many wives does your partner

have?

______________

98 = Don’t know

Q17: Does your partner have any partners/lovers in

addition to yourself?

1= Yes

0 = No >>>>> Q19

98 = Don’t know >>>>> Q19

Q18: How many partners/lovers does your partner have? ______________

98 = Don’t know

312

THIS INCLUDES THE RESPONDENT IF S/HE IS

IN A RELATIONSHIP

Q19: In general, do you and your current partner discuss

the following topics together:

Yes No Can’t say or

don’t want to

say

a) Things that have happened to your partner

during the day

1 0 3

b) Things that happen to you during the day 1 0 3

c) Your worries or feelings 1 0 3

d) your partner’s worries or feelings 1 0 3

Q20: In your relationship with

your current partner, how

often would you say that

you quarrelled?

1 = Less than once per month

2 = Once per month


4 = Once a week



7 = Every day

1 = Yes, all

the time

2 = Yes,

sometimes

3 =

No,

never

Q21: In your relationship with your current recent partner,

has your partner ever gotten angry if you speak with

another man/woman?

Q22: In your relationship with your current partner, have

you ever gotten angry if your partner speaks with

another man/woman?

Q23: In your relationship with your current partner, does

your partner ever suspect that you are unfaithful, and

not trust you?

1 = Very often

2 = A few times

3 = Rarely

4 = Never

Q24: In your relationship with your current partner, do you

ever suspect that your partner is unfaithful, and not

trust him/her?

1 = Very often

2 = A few times

3 = Rarely

4 = Never

Q25: Are you currently living together? 1 = Yes >>>>> Q30

0 = No

Q26: Why not? 1 = I am not the first wife

2 = S/he is travelling for

work or education

3 = Too early in the

relationship

4 = Waiting to be married

5 = Need a bigger house

6 = In this part of Ghana, we don’t have to

live with our partner when we are married

7 = S/he is migrating for work or education

8 = I have migrated here for work or

education

9 = We have separated

10 = Other

If not, can I

ask you some

questions

about your

partner?

Q1: How old is

your partner?

Q2: What is the

highest level of

education that

your partner

has received?

1 = Pre-school

2 = Primary

3 = Middle/JSS/JHS


5 = TECH/VOC

6 = Higher

98 = Don’t know

Q3: What would

best describe

your partner’s

1 = Self-employed

2 = Employee

3 = Casual work

7 = Apprentice

8 = Retired/too old to

work

313

current work

situation?

4 = Not working and not

looking

5 = Not working and looking

for work

6 = Student

9 = Disabled/sick

10 = Too young to work

98 = Don’t know

Q4: Are you currently married? 1 = Yes

0 = No >>>>>>> Q50

Q5: How many years have you been

married?

Q6: Which type of marriage have you

had?

1 = Customary marriage

2 = Mohammedan/Islamic

3 = Ordinance

I would like to ask you a few questions about what would happen if you were to divorce. I know

this might not be happening to you now, but your opinion will help us better understand family

life in Ghana.

Q7: Who would be likely

to get custody of your

children (0–4 years)

in the case of a

divorce?

1 = Myself

2 = My spouse

3 = Other family

members

4 = Children

choose

5 = Divided in

half between

husband and

wife

6 = Girls will go to the mother and boys to the

father

7 = Divided depending on other assets

8 = Depends on who is at fault or who initiated

the divorce

66 = Other

98 = Don’t know


to get custody of your

children (5–14 years)

in the case of a

divorce?

1 = Myself

2 = My spouse

3 = Other family

members

4 = Children

choose

5 = Divided in

half between

husband and

wife

6 = Girls will go to the mother and boys to the

father

7 = Divided depending on other assets

8 = Depends on who is at fault or who initiated

the divorce

66 = Other

98 = Don’t know

Q9: Would you expect the

main breadwinner to

provide financial

support for the

children in case of a

divorce?

1 = Yes

2 = Yes, but only if they

have custody of the

children

3 = Depends on who is at

fault or who initiated the

divorce

4 = Depends on the financial situation

0 = No

66 = Other

98 = Don’t know


main breadwinner to

provide financial

support for his/her

1 = Yes

2 = Yes, but only if they

have custody of the

children

4 = Depends on the financial situation

0 = No

66 = Other

98 = Don’t know

314

spouse in case of a

divorce?

3 = Depends on who is at

fault or who initiated the

divorce


to get immobile

assets, such as the

house and land, in

the case of a divorce?

1 = Myself

2 = My spouse

3 = Other family

members

4 = Divided in half

between husband

and wife

5 = Divided

depending on

custody of children

6 = Husband and wife get what belongs to

them

7 = Husband and wife get what they brought

into marriage

8 = Depends on who is at fault or who

initiated the divorce

66 = Other

98 = Don’t know


to get mobile assets,

such as vehicles,

furniture, savings,

livestock, in the case

of a divorce?

1 = Myself

2 = My spouse

3 = Other family

members

4 = Divided in half

between husband

and wife

5 = Divided

depending on

custody of children

6 = Husband and wife get what belongs to

them

7 = Husband and wife get what they brought

into marriage

8 = Depends on who is at fault or who

initiated the divorce

66 = Other

98 = Don’t know

Q13: For men: Did your family pay a

bride price?

1 = Yes >>> Q41

0 = No

Q14: For men: Why not?

Then next module

1 = We could not

afford it

2 = We do not follow

this system

3 = We are still in the process of buying the

items

4 = Other

Q15: For men: What did the bride

price/dowry include?

1 = Cash

2 = Livestock

3 = Land

4 = Cloth/clothes

5 = Drinks

6 = Jewellery

7 = Suitcase

8 = Furniture

9 = Bible or Quran

10 = Other gifts

Q16: For men: Did your partner

contribute to the bride price?

1 = Yes

2 = No >>>> Then next module

Q17: For men: What did your partner

contribute?

1 = Cash

2 = Livestock

3 = Land

4 = Cloth/clothes

5 = Drinks

6 = Jewellery

7 = Suitcase

8 = Furniture

9 = Bible or Quran

10 = Other gifts

Q18: For women: Did your family

receive a bride price on your

behalf?

1 = Yes >>> Q46

0 = No

Q19: For women: Why

not?

Then next module

1 = My partner’s family

could not afford it

3 = My partner’s family are still in the

process of buying it

4 = Other

315

2 = We do not follow this

system

Q20: For women: What did the bride


1 = Cash

2 = Livestock

3 = Land

4 = Cloth/clothes

5 = Drinks

6 = Jewellery

7 = Suitcase

8 = Furniture

9 = Bible or Quran

10 = Other gifts

Q21: For women: Did you contribute to

the bride price?

1 = Yes

2 = No >>>> Next module

Q22: For women: What did you

contribute?

Then next module

1 = Cash

2 = Livestock

3 = Land

4 = Cloth/clothes

5 = Drinks

6 = Jewellery

7 = Suitcase

8 = Furniture

9 = Bible or Quran

10 = Other gifts

Q23: For women whose husband has

more than one wife: Are you the

first, second, third wife?

RANK _______

Q24: Are you engaged? 1 = Yes

0 = No

Q25: Are you in the process of a

customary marriage?

1 = Yes

0 = No

For those who are currently not in a relationship:

Q1: If you are not currently in a relationship,

have you had a relationship in the past?

IF RESPONDENT HAS HAD MORE

THAN ONE RELATIONSHIPS IN THE

PAST, PLEASE REFER TO THE MOST

RECENT ONE

1 = Yes, but we are now separated/broke up

2 = Yes, but s/he left/disappeared

3 = Yes, but we are now divorced

4 = Yes, but I am now widowed

0 = No >>>>>> NEXT MODULE

Q2: If 4 to Q52, were you ever asked to marry

the brother/sister of your spouse who

passed away?

1 = Yes

0 = No

Q3: If 1 to Q53, did you agree to marry a

brother /sister of your spouse who passed

away?

1 = Yes

0 = No

Q4: After your spouse passed

away, were your spouse's

assets (land, property,

money…) passed on to you?

1 = Yes, it was all passed on to

me

2 = Yes, some of it was passed

on to me

3 = No, it was given to our

children

4 = No, it was given to

my spouse’s family

5 = No, it was given to

someone else

I would like to ask you some questions about your most recent relationship.

Q5: Were you in a relationship with more

than one person?

1 = Yes

0 = No >>>>>> Q60

Q6: If yes, how many persons were you in a

relationship with?

__________________ number

316

THIS INCLUDES ALL RELATIONSHIPS,

INCLUDING THE PRIMARY

RELATIONSHIP

Q7: Were you married to any of these other

persons?

1 = Yes

0 = No >>>>>> Q60

Q8: If yes, how many of these persons were

you married to?

THIS INCLUDES ALL MARRIAGES,

INCLUDING THE PRIMARY

MARRIAGE

__________________ number

For the remainder of these questions I would like to ask you about your most recent

primary/main partner. This could be the person that you live with.

Q9: Were you in a relationship with a man or

a woman?

1 = Man

2 = Woman

Q10: How old were you when you entered this

relationship?

___________ years

Q11: Was this an approximation? 1 = Yes

0 = No

Q12: How old were you when you ended this

relationship?

____________ years

Q13: Was this an approximation 1 = Yes

0 = No

I would like to ask you some questions about your most recent partner.

Q14: How old was your partner when you

started your relationship?

Q15: How old was your partner when you

ended your relationship?

Q16: What was the highest level of

education that your partner

had received?

1 = Pre-school

2 = Primary

3 = Middle/JSS/JHS


5 = TECH/VOC

6 = Higher

98 = Don’t know

Q17: What would best describe

your most recent partner’s

work situation at the time?

1 = Self-employed

2 = Employee

3 = Casual work

4 = Not working and not

looking

5 = Not working and

looking for work

6 = Student

7 = Apprentice

8 = Retired/too old to work

9 = Disabled/sick

10 =Too young to work

98 = Don’t know

Q18: How did you get to know your most

recent partner?

1 = Arranged meeting

2 = We are family (e.g. cousins)

3 =

Friends/classmates/colleagues/neighbours/cus

tomer

4 = At party, festival or other social event

5 = Other

Q19: Who influenced the decision

of choosing your partner?

1 = Myself

2 = Parents

3 = Extended family

4 = Church/Mosque/cult/sect

5 = Other

If respondent is man skip to Q 73

317

MULTIPLE ANSWERS

ALLOWED

Q20: For Women: Did your partner have any wives

in addition to your relationship?

1 = Yes

0 = No >>>>> Q73

98 = Don’t know >>>>> Q73

Q21: For Women: How many wives did your

partner have?

THIS INCLUDES ALL MARRIAGES, EVEN

THE MARRIAGE WITH THE RESPONDENT

______________

98 = Don’t know

Q22: Did your partner have any partners/lovers in

addition to yourself?

1= Yes

0 = No >>>>> Q75

98 = Don’t know >>>>> Q75

Q23: How many partners/lovers did your partner

have?

THIS INCLUDES ALL RELATIONSHIPS,

EVEN THE RELATIONSHIP WITH THE

RESPONDENT

______________

98 = Don’t know

Q24: In general, did you and your most recent partner discuss

the following topics together:

Yes No Can’t say or

don’t want to

say

a) Things that had happened to your partner during the

day

1 0 3

b) Things that had happened to you during the day 1 0 3

c) Your worries or feelings 1 0 3

d) Your partner’s worries or feelings 1 0 3

Q25: In your relationship with

your most recent partner,

how often would you say

that you quarrelled?

1 = Less than once per month (once

in a while)

2 = Once per month


4 = Once a week



7 = Every day

1 = Yes, all

the time

2 = Yes,

sometimes

3 = No,

never

Q26: In your relationship with your most recent partner,

did your partner ever get angry if you spoke with

another man/woman?


did you ever get angry if your partner spoke with

another man/woman?


did your partner ever suspect that you were

unfaithful, and not trust you?

1 = Very often

2 = A few times

3 = Rarely

4 = Never


did you ever suspect that your partner was

unfaithful, and not trust him/her?

1 = Very often

2 = A few times

3 = Rarely

4 = Never

Q30: Were you living together with your most

recent partner?

1 = Yes >>>>> Q83

0 = No

318

Q31: Why

not?

1 = I was not the first wife

2 = S/he was travelling for work or

education

3 = Too early in the relationship

4 = We were waiting to be married

5 = Needed a bigger house

6 = In this part of Ghana, we don’t have

to live with our partner when we are

married

7 = S/he was migrating for work or

education

8 = I had migrated for work or education

9 = We had separated

10 = Other

Q32: Were you married to your most recent

partner?

1 = Yes

0 = No >>> Q103

Q33: How many years were you married for?

Q34: Which type of marriage did you have? 1 = Customary marriage

2 = Mohammedan/Islamic

3 = Ordinance

I would like to ask you a few questions about what would happen if you were to divorce or if

you have had a divorce. I know this might not have happened to you, but your opinion will help

us better understand family life in Ghana.

Q35: Who would be likely to

get custody (or who got

custody) of your children

(0–4 years) in the case of a

divorce?

1 = Myself

2 = My ex-spouse

3 = Other family

members

4 = Children choose

5 = Divided in half

between husband and

wife

6 = Girls will go to the mother and

boys to the father

7 = Divided depending on other

assets

8 = Depends on who is at fault or

who initiated the divorce

66 = Other

98 = Don’t know


get custody (or who got

custody) of your children

(5–14 years) in the case of

a divorce?

1 = Myself

2 = My ex-spouse

3 = Other family

members

4 = Children choose

5 = Divided in half

between husband and

wife

6 = Girls will go to the mother and

boys to the father

7 = Divided depending on other

assets



66 = Other

98 = Don’t know


main breadwinner to

provide financial support

for the children, or did the

main breadwinner


for the children, in case of

a divorce?

1 = Yes

2 = Yes, but only if they have

custody of the children



4 = Depends on the

financial situation

0 = No

66 = Other

98 = Don’t know

319


main breadwinner to


for his/her spouse, or did

the main breadwinner


for his/her spouse, in case

of a divorce?

1 = Yes

2 = Yes, but only if they have

custody of the children



4 = Depends on the

financial situation

0 = No

66 = Other

98 = Don’t know


get (or who got) immobile

assets, such as the house

and land, in the case of a

divorce?

1 = Myself

2 = My ex-spouse

3 = Other family members

4 = Divided in half between

husband and wife

5 = Divided depending on

custody of children

6 = Husband and wife get what

belongs to them


they brought into marriage

8 = Depends on who is at fault

or who initiated the divorce

66 = Other

98 = Don’t know


get (or who got) mobile

assets, such as vehicles,

furniture, savings,

livestock in the case of a

divorce?

1 = Myself

2 = My ex-spouse

3 = Other family members

4 = Divided in half between

husband and wife

5 = Divided depending on

custody of children


belongs to them


they brought into marriage

8 = Depends on who is at fault

or who initiated the divorce

66 = Other

98 = Don’t know

Q41: For men: Did your family pay a bride price? 1 = Yes >>> Q94 0 = No

Q42: For men: Why not?

Then next module

1 = We could not afford it

2 = We do not follow this

system

3 = We were still in the process of

buying it

4 = Other

For men: What did

the bride

price/dowry

include?

1 = Cash

2 = Livestock

3 = Land

4 = Cloth/clothes

5 = Drinks

6 = Jewellery

7 = Suitcase

8 = Furniture

9 = Bible or Quran

10 = Other gifts

Q1: For men: Did your partner contribute to

the bride price?

1 = Yes


Q2: For men: What did your partner

contribute?

Then next module

1 = Cash

2 = Livestock

3 = Land

4 = Cloth/clothes

5 = Drinks

6 = Jewellery

7 = Suitcase

8 = Furniture

9 = Bible or Quran

10 = Other gifts

Q3: For women: Did your family receive a

bride price/dowry on your behalf?

1 = Yes >>> Q99

0 = No

Q4: For women: Why

not?

Then next module

1 = My partner’s family could

not afford it

2 = We do not follow this system

3 = My partner’s family were still

in the process of buying it

4 = Other

320

Q5: For women: What did the bride


1 = Cash

2 = Livestock

3 = Land

4 = Cloth/clothes

5 = Drinks

6 = Jewellery

7 = Suitcase

8 = Furniture

9 = Bible or Quran

10 = Other gifts

Q6: For women: Did you contribute to the

bride price?

1 = Yes


Q7: For women: What did you contribute?

Then next module

1 = Cash

2 = Livestock

3 = Land

4 = Cloth/clothes

5 = Drinks

6 = Jewellery

7 = Suitcase

8 = Furniture

9 = Bible or Quran

10 = Other gifts

Q8: For women whose husband has more

than one wife: Were you the first, second,

third wife?

RANK _______

Q9: Were you engaged to your most recent

partner?

1 = Yes

0 = No

Q10: Were you in the process of a customary

marriage with your most recent partner?

1 = Yes

0 = No

321

Reproductive decisions

I would like to ask you some questions about your sexual health and life. Please remember that

all of this information will be kept confidential, and your name will be stored separately to the

rest of the information that we gather.

Q1: How old were you when you first had

sexual intercourse? _____________

AGE

99 = no answer

88 = no sexual experience

>>>> Q17

Q2: Did you give consent to your first

sexual partner?

1 = Yes

0 = No

Is the respondent in a relationship?

ENUMERATOR TO CHECK

RELATIONSHIP STATUS

1= Yes

0 = No >>>>> Q19

For those in a relationship: Now we would like to ask you a few questions regarding sexual and

reproductive health.

Q3: For women: Are you pregnant at the moment? 1 = Yes

0 = No >>>>> Q12

Q4: Before you were pregnant did you and your partner

use birth control (methods to delay or avoid

pregnancy?)

1 = Yes

0 = No >>>>> Q7

Q5: Who made this decision?

1 = Yourself

2 = Your partner

3 = Self and partner

4 = Someone else

Q6: What type of birth

control did you and

your partner use?

(MULTIPLE ANSWER

ALLOWED)

FOR ALL, GO TO Q12

AFTER

1 = Condom

2 = Female condom

3 = Hormonal pill

4 = Hormonal implant

5 = IUD (coil)

6 = Sterilisation

7 = Abstinence

8 = Withdrawal

9 = LAM

10 = Foam/jelly

11 = Diaphragm

12 = Rhythm

13 = Emergency contraception

14 = Injectable

15 = Other

Q7: Why did you and your

partner not use birth

control?

(MULTIPLE ANSWER

ALLOWED)

FOR ALL, GO TO Q12

AFTER

1 = Partner didn’t allow

2 = Someone else did not

allow

3 = Makes me feel weak/ill

or other side effects

4 = Makes sex less

pleasurable

5 = Didn’t feel the need to

6 = Religious beliefs

7 = It is expensive

8 = I felt embarrassed

9 = It is not available

10 = We are trying to have

children

11 = Me/partner post-

menopausal

12 = Homosexual relationship

13 = Me/partner is infertile

14 = Tubes are tied,

reproductive organs removed

or vasectomy

15 = We had not heard of

contraceptive methods

16 = We don’t have sex

17 = I was not menstruating

18 = Other

Q8: Do you and your partner use birth control (methods

to delay or avoid pregnancy?)

1 = Yes

0 = No >>>>> Q11

322

Q9: Who made this decision?

1 = Yourself

2 = Your partner


4 = Someone else

Q10: What type of birth

control do you and your

partner use?

(MULTIPLE ANSWER

ALLOWED)

FOR ALL, GO TO Q12

AFTER

1 = Condom

2 = Female condom

3 = Hormonal pill

4 = Hormonal implant

5 = IUD (coil)

6 = Sterilisation

7 = Abstinence

8 = Withdrawal

9 = LAM

10 = Foam/jelly

11 = Diaphragm

12 = Rhythm

13 = Emergency contraception

14 = Injectable

15 = Other

Q11: Why are you and your

partner not using birth

control?

(MULTIPLE ANSWER

ALLOWED)



allow

3 = Makes me feel

weak/ill or other side

effects

4 = Makes sex less

pleasurable



7 = It is expensive




children

11 = Me/partner post-

menopausal

12 = Homosexual relationship

13 = Me/partner is infertile

14 = Tubes are tied,

reproductive organs removed

or vasectomy

15 = We have not heard of

contraceptive methods


17 = I am not menstruating

18 = Other

Q12: Do you and your partner use a safe sex method to

avoid sexually transmitted diseases and infections?

1 = Yes

0 = No >>>>> Q14

Q13: Which method do you and your partner use?

(MULTIPLE ANSWER ALLOWED)

FOR ALL, GO TO Q15 AFTER

1 = Condom

2 = Female condom

3 = Abstinence

4 = Other

Q14: Why don’t you and your

partner use any method?

(MULTIPLE ANSWER

ALLOWED)



allow

3 = Makes me feel

weak/ill or other side

effects

4 = Makes sex less

pleasurable


because we trust each

other


7 = It is expensive




children

11 = We have not heard of safe

sex methods


13 = Other

Q15: Over the last six months, or since you

met your partner, how often do you and

your partner engage in sexual

intercourse?

1 = Never

2 = Every day

3 = Several times a week

4 = Once a week

5 = Several times a month

323

6 = Occasionally

Q16: Do you think this is adequate for you? 1 = Yes

2 = No, it should be more

3 = No, it should be less

Q17: Has your partner ever refused to have

sexual intercourse with you for long

periods of time?

1 = Yes

0 = No

Q18: Who makes decisions about how many

children to have?

1 = Yourself

2 = Your partner


4 = Someone else

5 = We have not discussed this (it is too

early)

6 = We cannot have children

Q19: Have you ever heard of female

circumcision?

1 = Yes >>>> if woman skip to q 21. If man

skip to next module

0 = No

Q20: In a number of countries, there is a

practice in which a girl may have part of

her genitals cut. Have you ever heard

about this practice?

1 = Yes >>>> If man skip to next module

0 = No >>>> Skip to next module

For women:

Q21: Have you yourself ever had your

genitals cut?

1 = Yes


Q22: Could you please tell me what was done

to you at this time.

Was any flesh removed from the genital

area?

1 = Yes

0 = No

98 = Don’t know

Was the genital area just nicked without

removing any flesh?

1 = Yes

0 = No

98 = Don’t know

Was your genital area sewn closed? 1 = Yes

0 = No

How old were you when this occurred? Age in completed years ____________

95 = During infancy

98 = Don’t know

Who cut (or nicked) the genitals? 1 = Traditional circumciser

2 = Traditional birth attendant

3 = Other traditional person

4 = Doctor

5 = Trained nurse/midwife

6 = Other health professional

7 = Other

98 = Don’t know

324

5. Sexual Violence

Please can you tell me if

anyone has ever done or

tried to do the following

things to you:



household, family, work

colleagues etc.

a) M

ade

inap

pro

pri

ate

sex

ual

com

men

ts t

o y

ou

th

at m

ade

yo

u f

eel

un

com

fort

able

b)

To

uch

ed y

ou

in

an

in

app

rop

riat

e

and

sex

ual

way

th

at m

ade

yo

u f

eel

un

com

fort

able

c) P

hy

sica

lly

fo

rced

yo

u t

o h

ave

sex

ual

inte

rco

urs

e o

r p

erfo

rm a

sex

ual

act

wh

en y

ou

did

no

t w

ant

to?

d)

Oth

erw

ise

forc

ed y

ou

to

hav

e se

xu

al

inte

rco

urs

e o

r p

erfo

rm a

sex

ual

act

wh

en y

ou

did

no

t w

ant

to?

Th

is c

ou

ld

be

by

bla

ckm

aili

ng

, th

reat

enin

g o

r

scar

ing

yo

u.

e) H

ad s

exu

al i

nte

rco

urs

e o

r o

ther

sex

ual

act

wit

ho

ut

yo

u b

ein

g a

ble

to

giv

e y

ou

r p

erm

issi

on

?

f) N

ot

use

d p

rote

ctio

n e

ven

aft

er y

ou

ask

ed?

g)

A s

exu

al p

artn

er d

id n

ot

rev

eal

to

yo

u t

hat

he/

she

had

HIV

(an

d t

hey

kn

ew a

bo

ut

it)?

h)

Had

sex

ual

in

terc

ou

rse

or

per

form

ed

a se

xu

al a

ct w

ith

so

meo

ne

bec

ause

yo

u

felt

lik

e y

ou

did

no

t h

ave

a ch

oic

e o

r

yo

u w

ere

wo

rrie

d a

bo

ut

the

reac

tio

n.

i) P

enet

rate

d y

ou

wit

h a

n o

bje

ct a

gai

nst

yo

ur

wil

l?

Q1: 1 = Yes

2 = Someone tried >>>

NEXT ITEM

0 = No >>> NEXT

ITEM

Q2: Did this happen when

you were younger

than 15 years old?

1 = Yes, 0 = No

Q3: Has this happened in

the last 12 months?

1 = Yes,

0 = No >>> NEXT

ITEM

325

Q4: In the last 12 months,

was the person who

did this to you from

your current

household? 1 = Yes all

of them, 2 = Some of

them, 0 = No >>Q6

Q5: Who did this to you

from your current

household? Use IDs

from Household roster

Then skip to Q8

Q6: Who did this to you?

Use codes above

Q7: What was the sex of

this person?


Q8: I am sorry to ask

details about this

experience, but could

you tell me how they

physically forced you?


months, how often has

this happened? Use

codes above

Q10: Did this result in any

injuries?

1 = Yes, 0 = No >>>>>

Q12

326

Q11: What type of injury?

Q12: How has this affected

your daily activities?

USE CODES BELOW

Q13: If 6 to Q12: How many

days were you not able

to work over the last 12

months?


days were you not able

to go to school over the

last 12 months?

Q15: When you were a child, did you ever observe this kind of behaviour between your

parents or other members of your household?

1 = Yes, parents


0 = No

Q16: Think of 10 of your closest neighbours, how many of them do you know, or think,

experienced this kind of behaviour over the last 12 months? 98 = Don’t know

_______________ out of 10


THIS HAPPENED?

1 = Partner

2 = Former partner


4 = Parents

5 = Sons, daughters


7 = Uncles, aunts

8 = Grandparents

9 = Cousins, nieces, nephews, other

relatives


11 = Other relatives who do not live in this household

but in the community


13 = Friend

14 = Teacher

15 = Employer

16 = Classmates


18 = Police



1 = Once only

2 = Once in a while

3 = Less than once per month

4 = Once per month


6 = Once a week

7 = 2–3 times a

week

8 = Almost every

day

9 = Every day

327

CODES FOR Q11: WHAT TYPE OF INJURY CODES FOR Q12: HOW HAS THIS AFFECTED YOU?

(select multiple)

CODES FOR Q8: HOW

WERE YOU

PHYSICALLY FORCED?

1 = Non-genital injuries

(Cuts, punctures, bites,

Scratches, abrasion, bruises)

2 = Genital injuries (bruising,

inflammation, tenderness,

abrasions, or lacerations)

3 = Chronic pain

4 = Gastrointestinal

disorders

5 = Gynaecological

complications

6 = Migraines and other

frequent headaches

7 = Sexually transmitted

infections

8 = Cervical cancer

9 = Loss of pregnancy

10 = Premature birth

0 = Not at all

1 = Unable to concentrate at

work


school


domestic work

4 = Could not take care of the

children

5 = I had to miss work

6 = I could not go to

school

7 = I couldn’t do

domestic work

8 = Lost confidence in

own ability

9 = Left work

10 = Left school

11 = Live in constant

fear

12 = Other

1 = Pinning you down, so

that you could not get up

2 = Beating you up if you

tried to move

3 = Beating you until you

became unconscious

4 = Otherwise hurting you

so that you would not

move

328

Q17: Over the last 12 months, how many times

were you injured due to sexual violence?

1 = Once/twice

2 = 3–5 times

3 = Many (more than 5) times

Q18: Over the last 12 months, were you ever hurt

badly enough due to sexual violence that

you needed health care?

IF YES: How many times?

1 = Times needed health care:

________

2 = Yes, but don’t know

3 = Not needed >>> NEXT MODULE

Q19: Did anyone stop you from trying to receive

health care for these injuries?

1 = Yes

0 = No >>>>> Q24

Q20: Was the person who stopped you from your

household?

1 = Yes

0 = No >>>>> Q22

Q21: Who stopped you from your household? Use Household IDs

Q22: Who stopped you from outside your

household?

Use perpetrator codes

Q23: How many times has someone stopped you

from receiving health care over the last 12

months?

__________________

Q24: Over the last 12 months, have you ever

received health care these injuries?

IF YES: All of the time, or sometimes?

1 = Times needed health care:

________



Q25: How much did you spend on health care for

your injuries including medicines, fees and

transport (IF MULTIPLE VISITS: during

your last health visit)?

Q26: For your injury, did you have to spend any

nights in any health facility over the last 12

months?

IF YES: How many nights (in total)?

Number of nights in health facility:

________

If none, enter ‘00’

Q27: During your last health care visit, did you

tell a health worker the real cause of your

injury?

1 = Yes >>> NEXT MODULE

0 = No

Q28: During your last health care visit, why did

you decide not to tell the real cause of your

injury?


1 = I was afraid they would tell someone

2 = I was afraid of repercussions from the

person who hurt me

3 = I was afraid they would not believe me

4 = I was afraid they would not care

5 = I was afraid they would treat me badly

6 = Some of the people who work there

know me or my family

7 = Other

329

If respondent has reported no sexual violence, Read out loud: I know that these questions

are very personal and some people would rather not reveal this sensitive information to a

stranger. So we are giving everyone, no matter what they said before, the chance to reveal

this information privately and discreetly without me watching you. No matter what you

answered before, I am going to give you the questionnaire, and let you choose your

responses without me looking. You can turn the page over as soon as you are done

Q29: Read to respondent: Before you were 15 years old, had

anyone ever touched you sexually or made you do

something sexual that you did not want to do?

Now hand questionnaire to respondent.

Q30: Now I would like you to answer the same question, but

when you were older than 15 years old.


330

If respondent has been injured (Yes to ANY IN Q10) READ: I would like to learn more about the injuries resulting from the sexual violence you

have experienced

Please can you tell me if you have

ever done or tried to do the

following things to someone else:


partner, members of your family,


a) Made

inappropri

ate sexual

comments

to someone

that made

them feel

uncomfort

able

b) Touched

someone in

an

inappropri

ate and

sexual way

that made

them feel

uncomfort

able

c)

Physically

forced

someone to

have

sexual

intercourse

or perform

a sexual act

when they

did not

want to?

d)

Otherwise

forced

someone to

have

sexual

intercourse

or perform

a sexual

act? This

could be

by

blackmaili

ng,

threatening

or scaring

them

e) Had

sexual

intercourse

or

performed

a sexual act

with

someone

who was

not able to

give their

permission

?

f) Not used

protection

even after

someone

asked?

g) Not

reveal to a

sexual

partner

that you

had HIV

(and you

knew

about it)?

h) Felt that

someone

had sexual

intercourse

or

performed

a sexual act

with you

because

you felt like

they did not

have a

choice or

were

worried

about your

reaction

i)

Penetrated

someone

with an

object

against

their will?

Q1: 1 = Yes

2 = I tried >>> NEXT ITEM


Q2: Has this happened in the

last 12 months? 1 = Yes,


Q3: Over the last 12 months,

how often has this

happened? Above


was the person you did this

331

to from your household?

1 =Yes, 0 = No >>>>Q36


who did you do this to

from your household? Use

Household IDs

Q6: Over the last 12 months

who did you do this to?

Use codes above


person?


332

Please can you tell me if anyone

has ever done, or tried to do, the




household, family, work colleagues

etc.

a) Slapped

you or

thrown

something

at you that

could hurt

you?

b) Pushed

you or

shoved

you?

c) Hit you

with

his/her fist

or with

something

else that

could hurt

you?

d) Kicked

you,

dragged

you or

beaten

you up?

e) Choked

or

strangled

you on

purpose?

f) Burnt

you on

purpose?

g) Used a

gun, knife

or other

weapon

against

you?

h) Poured

hazardous

chemicals

or

substance

s (e.g.

acid) on

you?

i) FOR

MEN

ONLY:

Kicked or

pulled

your

external

genitalia?

Q1: 1 = Yes

2 = Someone tried >>>

NEXT ITEM


Q2: Did this happen when you

were younger than 15

years old?

1 = Yes, 0 = No




Q4: In the last 12 months, was

the person who did this to

you from your current

household? 1 = Yes all of

them, 2 = Some of them,

0 = No >>Q6

Q5: In the last 12 months, who

did this to you from your

current household? Use

IDs from Household roster

Then skip to Q8

333

Q6: In the last 12 months, who

did this to you?

Use codes below


person?



how often has this

happened?

Use codes below

Q9: Over the last 12 months

did this result in any

injuries?

1 = Yes, 0 = No >>>>> Q11

Q10: What type of injury?


did this ever end in a fight

where you also hit back

and retaliated?

Clarify: to hurt the person

doing this to you? USE

CODES BELOW

Q12: How has this affected your

daily activities over the

last 12 months? USE

CODES BELOW


days were you not able to

work over the last 12

months?

334


days were you not able to

go to school over the last

12 months?

Q15: For women, if answered yes to any of the above: During the times you were hit,

were you ever pregnant?

1 = Yes, 0 = No >>> Q17

Q16: For women, if answered yes to any of the above: If yes, did something happen

to your pregnancy?

1 = Yes, I had a miscarriage

2 = Still birth

3 = Premature delivery

4 = Neurological

complications to foetus

5 = Neurological

complications to foetus

which include long term

disability

6 = Other options

0 = Nothing



1 = Yes, parents

2 = Yes, other

members

0 = No



98 = Don’t know

_______________ out of 10

CODES FOR Q6 and 37: WHO DID THIS TO YOU? AND WHO DID YOU DO THIS TO? CODES FOR Q8 and 34: HOW OFTEN

HAS THIS HAPPENED?

1 = Partner

2 = Former partner


4 = Parents

5 = Sons, daughters


7 = Uncles, aunts

8 = Grandparents

9 = Cousins, nieces, nephews,

other relatives


11 = Other relatives who do not live in this household but in the

community


13 = Friend

14 = Teacher

15 = Employer

16 = Classmates


18 = Police



1 = Once only

2 = Once in a

while

3 = Less than

once per month

4 = Once per

month

5 = 2–3 times per

month

6 = Once a week



9 = Every day

335

CODES FOR Q10: WHAT TYPE OF

INJURY

CODES FOR Q11: DID YOU

EVER FIGHT BACK?

CODES FOR Q12: HOW HAS THIS AFFECTED YOU? (mark all

that apply)

1 = Cuts, punctures,

bites

2 = Scratch,

abrasion, bruises

3 = Sprains,

dislocations

4 = Burns

5 = Penetrating

injury, deep cuts,

gashes

6 = Ear or eye injury

7 = Fractures, broken

bones

8 = Broken teeth

9 = Other

1 = Never

2 = Once or twice

3 = Several times

4 = Many times/most of the time

0 = Not at all

1 = Unable to concentrate at work

2 = Unable to concentrate at school


domestic work

4 = Could not take care of the

children

5 = I had to miss work

6 = I could not go to school

7 = I couldn’t do domestic

work

8 = Lost confidence in own

ability

9 = Left work

10 = Left school

11 = Live in constant fear

12 = Other

336

If respondent has been injured (Yes to ANY IN Q9) READ: I would like to learn more about the

injuries resulting from the physical violence you have experienced.

Q19: Over the last 12 months, how many

times were you injured due to

physical violence?

1 = Once/twice

2 = 3–5 times

3 = Many (more than 5) times

Q20: Over the last 12 months, were you

ever hurt badly enough due to

physical violence that you needed

health care?

IF YES: How many times?

1 = Times needed health care: ________



Q21: Did anyone stop you from trying to

receive health care for these injuries?

1 = Yes

0 = No >>>>> Q26

Q22: Was the person who stopped you


1 = Yes

0 = No >>>>> Q24

Q23: Who stopped you from your

household?

Use Household IDs

Q24: Who stopped you from outside your

household?

Use perpetrator codes

Q25: How many times has someone

stopped you from receiving health

care over the last 12 months?

__________________

Q26: Over the last 12 months, have you

ever received health care these

injuries?

IF YES: All of the time, or

sometimes?

1 = Times needed health care: ________



Q27: How much did you spend on health

care for your injuries including

medicines, fees and transport (IF

MULTIPLE VISITS: during your last

health visit)?

Q28: For your injury, did you have to

spend any nights in any health

facility over the last 12 months?

IF YES: How many nights (in total)?

Number of nights in health facility:

________

If none, enter ‘00’

Q29: During your last health care visit,

did you tell a health worker the real

cause of your injury?

1 = Yes >>> NEXT MODULE

0 = No

Q30: During your last health care visit,

why did you decide not to tell the

real cause of your injury?


1 = I was afraid they would tell someone

2 = I was afraid of repercussions from the person

who hurt me

3 = I was afraid they would not believe me

4 = I was afraid they would not care

5 = I was afraid they would treat me badly

6 = Some of the people who work there know me or

my family

7 = Other

337

If respondent has reported no physical violence, Read out loud: I know that these

questions are very personal and some people would rather not reveal this sensitive

information to a stranger. So we are giving everyone, no matter what they said before,

the chance to reveal this information privately and discreetly without me watching you.

No matter what you answered before, I am going to give you the questionnaire, and let

you choose your responses without me looking. You can turn the page over as soon as

you are done

Q31: Read to respondent: Has anyone ever hurt you

physically by doing any of the things that I have just

mentioned?


338

6. Physical Violence


ever done or tried to do the





a)

Slapped

or thrown

somethin

g at

someone

that could

hurt

them?

b) Pushed

or shoved

someone?

c) Hit

someone

with your

fist or

with

somethin

g that

could hurt

them?

d) Kicked,

dragged

or beaten

someone

up?

e) Choked

or

strangled

someone

on

purpose?

f) Burnt

someone

on

purpose?

g) Used a

gun, knife

or other

weapon

against

someone?

h) Poured

hazardous

chemicals

or

substance

s (e.g.

acid) on

someone?

i) Kicked

or pulled

someone’s

external

genitalia?

Q1: 1 = Yes







how often has this

happened? Use codes above

Q4: Over the last 12 months, was

the person you did this to


1 =Yes, 0 = No >>>>Q37


who did you do this to from

your household? Use

Household IDs

Q6: Over the last 12 months who

did you do this to?

Use codes above

339


person?


7. Agency and Life Control

Q1: Please imagine a nine-step ladder, where on the

bottom, the first step, are those who are totally unable

to change their lives, while on step 9, the highest step,

stand those who have full control over their own life.

On which step are you?

__________________

How much do you agree with the following

statements?

1 = Strongly

agree

2 = Somewhat

agree

3 =

Neutral

4 = Somewhat

disagree

5 = Strongly

disagree

Q2: In uncertain times, I usually expect the best

Q3: It’s easy for me to relax

Q4: If something can go wrong, it will

Q5: I’m always optimistic about my future

Q6: I enjoy my friends a lot

Q7: It’s important for me to keep busy

Q8: I hardly ever expect things to go my way

Q9: I don’t get upset too easily

Q10: I rarely count on good things happening to me

Q11: Overall, I expect more good things to happen to me

than bad

340


has ever (even when you were a

child) done, or tried to do, the



partner, members of your household,


a) Insulted,

belittled, or

humiliated you in

private or in front

of other people

b) Threatened to

abandon you,

ignored you or

treated you

indifferently

c) Done things to scare

or intimidate you on

purpose? (For example

by breaking something

in front of you)

d) Threatened to

use a gun, knife or

other weapon

against you?

e) Threatened to

hurt you or

someone you

care about with

something other

than a weapon?

This includes

threats to take

away children or

someone you

care about

Q1: 1 = Yes

2 = Someone tried >>> NEXT ITEM


Q2: Did this happen to you when you

were younger than 15?

1 = Yes, 0 = No

Q3: Has this happened in the last 12

months?

1 = Yes, 0 = No >>> NEXT ITEM

Q4: Did this behaviour make you feel

worthless or make you doubt

yourself?

1 = Yes, all the time

2 = Yes, sometimes

3 = No, not at all


person who did this to you from

your current household? 1 = Yes all

of them, 2 = Some of them, 0 = No

>>Q7

341

Q6: In the last 12 months, who did this

to you from your current

household? Use IDs from

Household roster

Then skip to Q9

Q7: In the last 12 months, who did this

to you? Use codes above

Q8: What was the sex of this person?


Q9: Over the last 12 months, how often

has this happened?

Use codes above



1 = Yes, parents


0 = No



98 = Don’t know

_______________ out of 10


THIS HAPPENED?

1 = Partner

2 = Former partner

3 = In-laws/partner’s

family

4 = Parents

5 = Sons, daughters


7 = Uncles, aunts

8 = Grandparents


other relatives


11 = Other relatives who do not

live in this household but in the

community

12 = Other relatives who live

outside this community

13 = Friend

14 = Teacher

15 = Employer

16 = Classmates


18 = Police


20 = Other people outside the

community

1 = Once only

2 = Once in a while


month

4 = Once per month


6 = Once a week


8 = Almost every

day

9 = Every day

342


ever done, or tried to do, the





a) Insulted,

belittled, or

humiliated

someone in

private or in front

of other people

b) Threatened to

abandon someone,

ignored someone or

treated them

indifferently

c) Done things to

scare or intimidate

someone on

purpose? (For

example by breaking

things in front of

someone)

d) Threatened to

use a gun, knife or

other weapon

against someone?

e) Threatened to

hurt someone or

someone they

care about with

something other

than a weapon?

This includes

threats to take

away children or

someone they

care about

Q1: 1 = Yes




months?



often has this happened?

Use codes above

Q4: Over the last 12 months, was the

person you did this to from your

household?

1 =Yes, 0 = No >>>>Q17

Q5: Over the last 12 months, who did

you do this to from your

household? Use Household IDs

Q6: Over the last 12 months who did

you do this to? Use codes above



343

8. Economic Violence

Q1: Do you have a bank account? 1 = Yes, 0 = No >>>>>>Q3

Q2: How much savings do you have in your bank account? _________________

Q3: Do you use mobile banking? 1 = Yes, 0 = No >>>>>Q5

Q4: How much savings do you have in your mobile banking account? _________________

Q5: Do you have any other form of savings? 1 = Yes, 0 = No >>>>>Q7

Q6: How much are these total savings? __________________


has ever (even when you were a

child) done, or tried to do, the



partner, members of your household,


a) Refused to

give you

enough chop

money even

though you

think he/she

has enough

money to

spend on other

things?

b) Taken

cash or

withdrawn

money from

your bank

account or

other savings

without

permission?

c)

Controlled

your own

belonging

s and/or

your

spending

decisions?

d)

Destroyed

or damaged

property

that you

have

material

interest in?

e)

Prohibited

you from

working or

forced you

to quit your

work?

f) Forced you

to work

against your

will?

g) Refused to

give you or

denied you food

or other basic

needs?

Q7: 1 = Yes

2 = Someone tried >>> NEXT

ITEM



months?



person who did this to you from

your current household? 1 = Yes

all of them, 2 = Some of them,

0 = No >>Q11

344


this to you from your current

household? Use IDs from

Household roster

Then skip to Q7


this to you? Use codes below




often has this happened?

Use codes above



1 = Yes, parents


0 = No



98 = Don’t know

_______________ out of 10


THIS HAPPENED?

1 = Partner

2 = Former partner

3 = In-laws/partner’s

family

4 = Parents

5 = Sons, daughters


7 = Uncles, aunts

8 = Grandparents


other relatives


11 = Other relatives who do not

live in this household but in

the community

12 = Other relatives who live

outside this community

13 = Friend

14 = Teacher

15 = Employer

16 = Classmates


18 = Police


20 = Other people outside the

community

1 = Once only

2 = Once in a while


month

4 = Once per month


6 = Once a week

7 = 2–3 times a

week

8 = Almost every

day

9 = Every day

345

Please can you tell me if you

have ever done, or tried to do,

the following things to someone

else:




a) Refused to

give someone

enough chop

money even

though you

have enough

money to

spend on

other things?

b) Taken

cash or

withdrawn

money from

someone’s

bank account

or other

savings

without

permission?

c)

Controlled

someone’s

belongings

and/or their

spending

decisions?

d) Destroyed

or damaged

property that

someone has

a material

interest in?

e) Prohibited

someone

from

working or

forced

someone to

quit their

work?

f) Forced

someone to

work against

their will?

g) Refused

to give or

denied

someone

food or other

basic needs?

Q16: 1 = Yes




months?



often has this happened? Use

codes above

Q19: Over the last 12 months, was the

person you did this to from your

household?

1 =Yes, 0 = No >>>>Q7

Q20: Over the last 12 months, who did

you do this to from your

household? Use Household IDs

Q21: Over the last 12 months who did

you do this to? Use codes above



346

9. Leaving Home

Q1: Over the last 12 months, have you ever left your home, even if

only for the night, because of a problem or situation at home?

1 = Yes

0 = No >>> NEXT

MODULE

Q2: Was this following any of the treatment or behaviour we have

previously discussed?

1 = Yes

0 = No >>>>Q4

Can you tell me which

treatment or behaviour in

particular?

Mark all that apply

Produce list of violence and behaviour that respondent

has said have happened to him/her over last 12 months

Q3: What were the reasons you left

the last time?

MARK ALL MENTIONED

0 = No particular reason

1 = Encouraged by friends/family

2 = Could not endure more

3 = Badly injured

4 = Afraid person/persons would kill me

5 = The person/persons who did this to me threatened or

tried to kill me

6 = The person/persons who did this to me threatened or

hit the children

7 = Saw that children are suffering

8 = Thrown out of the home

9 = Encouraged by organisation

10 = Encouraged by health workers/teachers

11 = Encouraged by police

12 = Other

Q4: How long did you stay away

the last time?

RECORD NUMBER OF

DAYS OR MONTHS

1 = Number of days (if less than one month)

_____________

2 = Number of months (if 1 month or more)

_____________

3 = Left or did not return to person who did this to me

>>>> Next module

Q5: Why did you return?

MARK ALL MENTIONED

1 = Didn’t want to leave children

2 = Sanctity of marriage

3 = For sake of family/children

4 = Couldn’t financially support children

5 = Loved person who did this to me

6 = Person who did this to me asked me to come back

7 = Encouraged to return by family

8 = Encouraged to return by health workers/teachers

9 = Encouraged to return by police

10 = Forgave person who did this to me

11 = Thought person who did this to me would change

12 = Person who did this to me threatened

partner/children

13 = Ran out of money

14 = Do not have another home to go to

15 = I was concerned about safety of people I was with

16 = I was concerned about safety of people left behind

347

17 = Other

10. Access to Services

Verify if respondent has ever experienced any violence? 1 = Yes

0 = No >>>>> Q8

Q1: After experiencing any of the types of violence

and treatment discussed previously, have you

ever tried to approach a person or organisation

for help or support?

1 = Yes >>>>>>>Q3

0 = No

Q2: Why not?

THEN GO TO

Q12

1 = Did not know who to ask

2 = Person/organisation too far

3 = Nobody can help me

4 = I don’t need help

5 = If my partner finds out it will

cause trouble

6 = If other family member finds out

it will cause trouble

7 = I don’t trust anyone

8 = Violence normal/not

serious

9 = Embarrassed/ashamed

10 = Afraid would not be

believed

11 = Afraid would be blamed

12 = Afraid would end

relationship

13 = Afraid would lose

children

14 = Bring bad name to family

15 = Other

Q3: What made you

approach

someone?

1 = Encouraged by friends/family

2 = Could not endure more

3 = Badly injured

4 = Afraid partner would kill me

5 = Partner threatened or tried to kill

me

6 = Partner threatened or hit

the children

7 = Saw that children are

suffering

8 = Thrown out of the home

9 = Encouraged by

organisation

10 = Other

Q4: Q5: Q6: Q7:

Who did you approach? How satisfied were

you with the way

that they dealt with

your case?

1 = Completely

satisfied

2 = Somewhat

satisfied

3 = Somewhat

dissatisfied

4 = Completely

dissatisfied

Did you have

to pay

anything to

this person or

organisation?

1 = Yes, I paid

a fee

2 = Yes, I paid

a bribe

0 = No >>>>

Next Item

How much

did you have

to pay?

A A family member or friend

B Health centre/hospital/other

health service provider

C Police

D Domestic Violence and Victim

Support Unit (DOVVSU)

348

E Shelter

F NGO, CSO or social worker

G Lawyer or member of the Court

H Legal aid group

I Queen mother of your

community

J Traditional leaders

K Religious leaders

L Religious group

M Community group

N Other

Thank you very much for answering these questions. We are almost at the end, and we have only

three more topics we would like to briefly discuss. First, we would like to ask you about access to

services in your area.

Q8: How long does it take you to get to your

nearest police station?

________ minutes

98 = Don’t know

Circle one:

1 = Foot

2 = Car

3 = Bicycle

4 = Motorcycle

5 = Tricycle

6 = Boat/canoe

7 = Public transport (e.g. bus or tro-tro)

8 = Truck


10 = Tractor

Q9: Does your nearest police station have a

domestic violence victim support unit

(DOVVSU)?

1 = Yes >>>>Q11

0 = No

98 = Don’t know

Q10: Do you know where the closest domestic

violence victim support unit (DOVVSU) is?

1=Yes

0=No

Q11: If not, how long would it take you to get to

the nearest domestic violence victims support

unit (DOVVSU)?

______________ minutes

98 = Don’t know

Circle one:

1 = Foot

2 = Car

3 = Bicycle

4 = Motorcycle

5 = Tricycle

6 = Boat/canoe

7 = Public transport (e.g. bus or tro-tro)

8 = Truck


10 = Tractor

349

How much would you trust the following if a friend

needed help after being physically assaulted in the home

by their partner?

1 = Completely trust

2 = Somewhat trust

3 = Somewhat distrust

4 = Completely distrust

98 = Never heard of them

Q12: Health centre/hospital/other health service provider

Q13: Police

Q14: Domestic Violence and Victim Support Unit (DOVVSU)

Q15: Shelter

Q16: NGO, CSO or social worker

Q17: Lawyer or member of the Court

Q18: Legal aid group

Q19: Queen mother of your community

Q20: Traditional leaders

Q21: Religious leaders

Q22: Religious group

Q23: Community group

Q24: Does Ghana have a law

against domestic violence?

1 = Yes

2 = Yes, but I do not

understand it

3 = Yes, but I do not know

what is in there exactly

4 = No >>>>>>NEXT

MODULE

98 = Don’t know

>>>>NEXT MODULE

Q25: If yes, where have you heard

about it?

1 = Friends/family

2 = Neighbours

3 = Community leaders

4 = Religious leaders

5 = Organisation

6 = Police

7 = School

8 = TV

9 = Radio

10 = Newspaper

11 = Books

12 = Other

11. Health

I would now like to ask you a few questions about your health and use of health services.

Q1: In general, would you describe your

overall health as …?

1 = Very good

2 = Good

3 = Fair

4 = Poor

5 = Very poor

Q2: Do you have any chronic conditions? 1 = Yes; 0 = No >>>>>Q4

Q3: Please indicate which

chronic condition(s)

you have?

MULTIPLE

ANSWERS

1 = Diabetes

2 = Asthma

3 = Lung disease

4 = Heart disease

5 = Arthritis or other

rheumatic disease

6 = Cancer

7 = High blood pressure

8 = Physical disability which restricts

certain activity

9 = Mental disability which restricts

certain activity

10 = Other chronic condition

Q4: Which of the following describes your

smoking habit?

1 = I used to smoke but not anymore

2 = I currently smoke >>> SKIP TO Q6

0 = I never smoked >>> SKIP TO Q7

1 = Daily 2 = Occasionally

Q5: Did you use to smoke daily or occasionally?

350

Q6: Do you smoke daily or occasionally?

Q7: Do you drink alcohol? 1 = Yes; 0 = No >>>>if in relationship Q9 if not Q11

Q8: How often do you drink

alcohol?

1 = Every day or nearly every day

2 = More than half the week but not

everyday

3 = Once or twice a week

4 = 1–3 times a month

5 = Less than once a

month

Q9: Does your partner drink alcohol? 1 = Yes, 0 = No >>>> Q11

Q10: How often does your

partner drink alcohol?

1 = Every day

2 = More than half the week but not

every day

3 = Once or twice a week

4 = 1–3 times a month

5 = Less than once a

month

Q11: Has anyone ever stopped you from or did not allow you to receive

health care when you needed it? Clarify: This could be by not allowing

respondent to visit health care facility or refusing to pay the fees or

costs of medicine.

1 = Yes

0 = No >>>>>

Q17

Q12: Has this happened over the last 12 months? 1 = Yes; 0 = No >>>> Q17

Q13: Was the person who stopped you from your household (over

the last 12 months)?

1 = Yes; 0 = No >>>>> Q15

Q14: Who stopped you from your household (over the last 12 months)? Use Household IDs

Q15: Who stopped

you from

outside your

household

(over the last

12 months)?

1 = Partner

2 = Former partner


4 = Parents

5 = Sons, daughters

6 = Uncles, aunts

7 = Grandparents


other relatives


10 = Other relatives who do not live in this

household but in the community

11 = Other relatives who live outside this

community

12 = Friend

13 = Teacher

14 = Employer


16 = Police



Q16: How many times has someone stopped you from receiving health care

over the last 12 months?

___________

Q17: In the last 30 days, have you felt sick or unwell? 1 = Yes; 0 = No >>>Q20

Q18: In the last 30 days, did you personally

consult a doctor or other professional

or health worker or traditional healer

because you were sick? Clarify: This

does not include health check-ups.

IF YES: Whom did you consult?


1 = No one

2 = Doctor

3 = Nurse

4 = Counsellor

5 = Traditional healer

6 = Pharmacist/chemical

seller

7 = Midwife

8 = Traditional birth

attendant

9 = Herbalist

10 = Spiritualist

11 = Prayer camp

12 = Other

Q19: Were you unable to work because of your sickness? IF YES: How many days? _________

During the last 30 days, about how

often did …

All the

time

Most

of the

time

Some

of the

time

A little

of the

time

None

of the

time

Q20: you feel nervous?

Q21: you feel hopeless?

Q22: you feel restless or fidgety?

Q23: you feel that everything was an effort?

351

Q24: you feel so depressed that nothing could

cheer you up?

Q25: you feel worthless?

Q26: has the thought of ending your life been

on your mind?

If there are children in the household aged 14 and younger

Adults use certain ways to teach children the right behaviour or to address a behaviour problem.

I will read various methods that are used and want you to tell me if you have used this method

with any children (aged 14 and younger) in your household in the past month.

In the last month, have you or

anyone else in your household

1 = Yes, 0 = No

Q1: Taking away privileges

Q2: Forbidding something that he/she liked

Q3: Not allow him/her to leave the house

Q4: Explaining why the behaviour was wrong

Q5: Giving him/her something else to do

Q6: Ignoring/refusing to communicate to the child

Q7: Shouting, yelling at or screaming at him/her

Q8: Calling him/her dumb, lazy or another name like that

Q9: Locking child up in room

Q10: Shaking him/her

Q11: Spanking, hitting, pushing or slapping him/her on the

bottom with bare hand

Q12: Hitting him/her on the bottom or elsewhere on the

body with something like a belt, comb, hairbrush, cane,

stick or other hard object

Q13: Hitting or slapping him/her on the face, head or ears

Q14: Hitting or slapping him/her on the hand, arm or leg

Q15: Beating him/her up, that is hitting him/her over and

over as hard as one could

Q16: Physically hurting them in any other way (e.g. burning)

12. Children

Finally, I would like to ask you a few questions about how the children of this household spend

their time.

Q17: How many hours do you spend with your

child/children (aged 14 and below) each day?

98 = Don’t know

99 = Not applicable

Q18: How many hours do you spend supervising your

child/children (aged 14 and below) during homework

each day? Should be less than Q17

98 = Don’t know

99 = Not applicable

Q19: How many hours does your child spend with other

adult household members?

98 = Don’t know

99 = Not applicable

Q20: Do you regularly attend Parent-Teacher Association

(PTA) meetings?

1 = Always

2 = Often

352

3 = Sometimes

4 = Rarely

5 = Never

99 = Not applicable

Q21: Does someone else from this household regularly

attend PTA meetings?

1 = Yes

0 = No

Thank you very much for your time! We know it is not easy to answer all these questions, and

we really appreciate that you took the time and patience. Since we have talked about issues

related to violence, we recommend that you do not speak to others about this part of the

questionnaire. Otherwise, some people in the community might get upset with you or others who

answered the survey. Are there any questions you have for us, or is there anything that you

would like to let us know?

_________________________________________________

353

Q3: Qualitative instruments

1. Focus Group Discussion Guideline – Domestic Violence Ghana 2015

Approach: For four separate and mixed FGDs with men and women. 1:30 hours /8–12

people

Group 1: Married/living together men (aged 18–50)

Group 2: Married /living together women (aged 18–50)

Group 3: Unmarried/youth (aged 18-25)

Group 4: Mixed (aged 50+)

Facilitator introduces the team, explains consent and the purpose

Thanks everyone for coming. Good morning/afternoon/evening and welcome. Thank you

all for taking the time to be part of this discussion. We very much appreciate your

willingness to participate.

My name is [name of facilitator], and I will serve as the moderator for today’s focus group

discussion. Assisting me is [name of assistant], who will take notes. This study is part of

the National Survey on Domestic Violence 2015 being conducted by the Institute of

Development Studies, University of Sussex (IDS) for the Ministry of Gender, Children

and Social Protection and DfiD, UK.

Participation in this research is voluntary. The discussion may last for about 45 to 90

minutes. All information provided will be kept secret. We will not ask your real name

and/or your address. You can withdraw from this interview anytime you want, or refuse

to answer any questions that you are not comfortable with.

The purpose of today’s discussion is to learn from you about:

The causes and consequences of domestic violence

Which services survivors of domestic violence use

What works well, what does not

How these services can be improved or strengthened

The findings will also inform civil society organisations and help them to reach out

to survivors of domestic violence and lobby for more programmatic interventions

and changes in laws.

There will be a report that will be published internationally.

We will be spending a few days in the community trying to find out about the different

roles of women and girls and men and boys in the community and in the family, and

discuss whether and how views and experiences with domestic violence has changed

compared to the past.

354

Facilitator explains the consent forms and the value of recording, and checks

permission.

Facilitator and Note-taker help participants complete consent forms.

Note-taker explains that we give people a number and use the numbers for the note-

taking.

Facilitator asks group for permission to record the discussion.

Rules

We would like you to feel safe here. We are here to learn from you; you have the knowledge

to help us understand these issues. We are not going to judge you and will listen to

everything you have to say. To do this, we would like to employ some guidelines for the

session to make sure we are safe and have a good discussion; the guidelines will apply to

all of us.

1. THERE ARE NO RIGHT OR WRONG ANSWERS

Every person’s experiences and opinions are important.

Speak up whether you agree or disagree.

It’s ok to have a different opinion; but everyone’s opinion matters equally.

We are interested to hear everybody’s opinion.

2. WE WOULD LIKE YOU TO DO THE TALKING.

Don’t feel like you have to respond to me all the time.

However, we would like everyone to participate.

So if you’re talking a lot, I may ask you to give others a chance.

And, I may call on you if I haven’t heard from you in a while.

3. WHAT IS SAID IN THIS ROOM STAYS HERE

Your experiences and input is very important and will feed into the research report – but

not what you say individually, unless you would like to be named.

We have name tents here in front of us today so that we can all know who’s sitting here

with us, but no names will be included in any reports!

Discussion starts.

Facilitators start with the questions. They do not read the headings.

Community’s perception of violence: What is violence?

1. Is it acceptable if a man beats his wife for disagreeing with him? (Allow different

opinions and allow participants to state reasons for their answers.)

If yes, why? …………………………………

If no, why? ……………………………

355

2. Is it acceptable if a woman beats her husband for having a child with someone

outside the family? (Allow different opinions and allow participants to state reasons

for their answers.)

If yes, why? …………………………………

If no, why? ……………………………

3. Is it acceptable if a man beats his wife for refusing to have sex with him? (Allow

different opinions and allow participants to state reasons for their answers.)

If yes, why? …………………………………

If no, why? ……………………………

4. Is it acceptable if a man refuses to give ‘chop money’ to his wife? (Allow different

opinions and allow participants to state reasons for their answers.)

If yes why? …………………………………

If no, why? ……………………………

5. Is it acceptable for men to tell women they cannot work?

If yes why? …………………………………

If no, why? ……………………………

6. Is it acceptable for men to beat children and youth in the home?

If yes why? …………………………………

If no, why? ……………………………

7. Is it acceptable for women to beat children and youth in the home?

If yes why? …………………………………

If no, why? ……………………………

8. Is it acceptable for fathers to stop their daughters from attending school?

If yes, why? ……………………………………………………..

If no, why? ………………………………………………………..

When should one intervene?

Allow different opinions and allow participants to state reasons for their answers.

9. What would you do if you were passing along a courtyard in your neighbourhood

and saw a man and woman shouting at each other?

10. What would you do if you saw the man now beating the woman?

11. What would you do if you saw a woman now beating the man?

12. What would you do if you saw the man/woman take out an object? What would

you do if you saw the woman/man now hurting the man/woman with the object?

Causes of domestic violence

13. Do we know whether there is disagreement and fights by people/men and women

from the same family in this community? If Yes, probe: [Who is involved in these

fights? What is the fighting about? How often do we hear or see these fights?]

14. So, why is there fighting within our families in this community

15. What are the various types of disagreement within families?

16. What typically happens during these fights/disagreements? (Describe these fights.)

356

17. Do men also experience violence in the family? What does this look like? How is

this different from how women experience violence in the domestic context?

18. Are weapons involved? (If yes, what type of weapons?)

19. Who is affected by these fights? How? Why? (Probe: children, elderly members,

others in family?)

20. What is the immediate impact? And what is the long-term impact on those affected?

Intervention and mediation

21. How do these domestic fights come to an end? Who do people go to for help? Who

is involved?

22. In your community, what kind of services are there to help people who are hurt by

their family? (Probe: list the services to understand the ‘in principle’ facets of

government support.)

23. Which of these services do people use the most? Why?

24. What has been your experience with government support for people hurt by other

people in their family? [Probe and discuss satisfaction with police handling of case

reporting/help at health centres.]

Social cost of domestic violence

25. Are others in the community affected by a family’s occasions of violence/blow-up?

Identify who. How are these persons affected? [Probe: different categories of

persons identified and how they are impacted by the violence.]

26. What are the costs to the rest of the community by a family’s experience of violence?

What would make violence acceptable to people/the community?

27. What would make members of the family turn a blind eye to violence within a

family?

28. What are the expectations of masculinity (breadwinner, disciplinarian) that can

make men feel they need to behave in a certain dominant or violent way?

29. What are the expectations of femininity (submissiveness) that can make women feel

they need to accept certain dominant or violent ways from men?

30. Discuss what are social pressures that can lead to people hurting one another in a

family or household – family and community pressure to encourage or discourage

prosecution.

31. What are the government pressures to encourage or discourage prosecution of

cases?

N.B. These are just guidelines. If any interesting additional issues emerge, the

discussion will also explore these. Facilitator asks participants if they have any

questions for us. Facilitator thanks participants for their time and the interesting

information they have shared.

357

2. Observation and In-Depth Interview Guideline on Domestic Violence Ghana:

Health Care Providers

Approach: Observations and in-depth interviews at health facilities (two per region and

>40 staff. Two staff per site). In each urban and rural site, health facilities that provide

emergency care for victims of domestic violence are visited. In each of these sites, both

observations and interviews should take place. Purposive sample of individuals who are

directly involved in domestic violence. These can be leaders or front-line staff or both

depending on the facility. Facilities are equally divided across selected urban and rural

sites.

Facilitator introduces the team, explains consent and the purpose.

Thanks for giving us your time. We very much appreciate your willingness to participate.

My name is [name of facilitator], and I will serve as the interviewer for today’s discussion.

Assisting me is [name of assistant], who will take notes. This study is part of the National

Survey on Domestic Violence 2015 being conducted by the Institute of Development

Studies, University of Sussex (IDS) for the Ministry of Gender, Children and Social

Protection and DfiD, UK. It covers all 10 regions of Ghana, urban and rural areas.





The purpose of today’s visit and discussion is to learn from you about:

The causes and consequences of domestic violence, and which services are available

to survivors of domestic violence



The findings will inform policy recommendations to the Ministry of Gender,

Children and Social Protection. It will also enable civil society organisations to reach

out more directly to survivors of domestic violence and lobby for more

programmatic interventions and changes in laws.

There will be a report that will be published.


permission.

Facilitator and Note-taker help to complete the consent form.

Note-taker fills out the clip-on sheet.

Facilitator asks for permission to record the discussion.

358

Note to Facilitator: You start with broad opening questions to help us understand some

more detail about the participant. It’s important to learn more about the person’s work;

this will help us to understand what they say, and what they might not feel comfortable

saying. It helps to understand the language that the participant will use.

Discussion starts.

Overview, background: tell me about yourself

1. Can you tell me about the work you do?

Probe:

a. As a health professional, what are some of your main responsibilities?

b. As a health professional, what are some of the main challenges in your work?

c. As a health professional, do you see any problems with the way men treat

women? With the way women treat men? With the way parents treat their

children?

Forms and causes of domestic violence

2. Can you tell me what you know about disagreement and fights by people from the

same family?

3. What kinds of fights do you see? [Probe to find about forms:

a. Are these fights verbal, physical, or sexually related?

b. Who is involved? Probe for: partners, parents and children, adults and non-

biological children?]

4. Why is there fighting/abuse and violence within families?

[Probe to find out about causes: Are they about money, about infidelity, about school

etc.?]

5. Based on your observations – as a professional – how often do you think it happens?

[Probe: At least once a week or once a month?]

6. Are there specific times when this fighting takes place?

[Probe: Any seasons or specific days? Is it linked to school terms? Is it linked to

monthly payment cycles?]

Impact of domestic violence

7. Who is affected by fights in the family? How? Why these people? [Probe: the range

of effects – they include losing confidence, being afraid to leave the house, not being

able to work and becoming trapped in a vicious cycle of dependence, for example.

How are these effects distributed across members of the household, who are

affected by the violence?]

8. Do men also experience violence in the family? What does this look like? How is this

different from how women experience violence in the domestic context?


[Probe for economic, psychological, health impacts.]

359

Acceptability of violence and social norms

10. What would make members of a family turn a blind eye to violence?





13. What would make health officers turn a blind eye to reported cases of domestic

violence? (i.e. discourage referral to police and other institutions for help)

14. Discuss – what are the social pressures – family and community pressure to

encourage or discourage health officers’ referral to the police for prosecution.


reported cases of domestic violence?


16. How do these fights come to an end? Who do people go to for help? Who is

involved? [Probe: to find out about government and non-government service

providers:

a. Are these government-appointed people (nurses, social workers, teachers)?

b. Do people feel more comfortable going to community leaders than to police

officers to get support?

c. Do people feel more comfortable going to certain police (men/women,

senior/junior) than other people/service providers? Why?

d. What kinds of things matter in how people make decisions to get support?

What catalyses these decisions?]

17. What kind of services are there to help people? [Probe to get a list of services to

understand the ‘in principle’ facets of government support as the participant

understands this.]

18. Which of these services do people use the most? Why? Which are not useful?


people in their family? (Discuss satisfaction with police handling of case

reporting/help at health centres.)

20. What do you think needs to happen to:

a. Help those people who commit violence to stop?

b. Help people who experience violence to get support?

c. Help government services reach the right people to end the violence?

d. Which services are seen to be most crucial? Why?

21. How many health cases from domestic violence did you received in this health

facility in:

2012 ……….………………………………………………………………

2013 ……….………………………………………………………………

2014 ……….………………………………………………………………

22. How many of these cases were female victims? How many were male victims

(Probe: ask for rough estimate on average/percentage.)

360

Availability of logistical support

23. Does the facility have specialised departments offer services to victims of violence

such as:

Rape-related services such as post-exposure prophylaxis for HIV

prevention (PEP), sexually transmitted infection (STI) referrals, and

emergency contraceptives (‘morning-after pills’)?

Legal referrals including prosecution advice for victims of violence (Is this

advice also given and tailored to men, and to gay men/women?)

Counselling services, mental health services

Any other (name service)………………………………………………………

Availability of protocols

24. Is domestic violence different from other forms of violence? Does the health unit

have formal protocols to follow when dealing with victims of domestic violence?

Yes

No

If Yes, name the documents …………………………………………………………..

If yes, what do these protocols deal with?

Emergency response to victims of violence

Priority for victims of abuse

Writing of medical reports

Referral to specialised units (paediatrician/social workers/police)

Capacity-building

25. Sensitising the health professionals on domestic violence. Has health centre staff

undergone:

Initial training on gender-based violence and on domestic violence

Continuing training on domestic violence

Specialised in listening skills for victims story

Training in writing medical reports

Affordable health care

26. Are medical reports free?

If No, How much is charged? …………………………………..

Is this a formal, required fee by the government or a personal fee charged by the

doctor? ……………………………………………………………………..

27. How are victims who are unable to pay fee for medical reports able to obtain the

report to forward to police for prosecution of the case?

Continuum of care

28. Does this health site work with any other service providers? If yes, which? (Police,

social workers) Name the organisations here

………………………………………………………………………………………..……..

361

29. How effective is this collaboration?

30. Can you give me an example where this site collaborated with other service

providers to give support to someone who had experienced domestic violence?

[Probe: Who was involved? What did each ‘sector’ do? What were the outcomes for

the person involved? Was the collaboration successful? What made it work well?]

OBSERVATIONS

While waiting look for:

Are files stored away or can the public see these easily? Do you hear staff talk about cases

and private affairs of patients in front of you when you are in a public area? Is there a

private room or other privacy measures for patients when they speak with medical staff?

Ask: Can you show us the building and the services as if we are a victim of domestic

violence who comes to report? Can you show us the process step by step? Where would

we go first, and what would we do? Who would help us? Where would we sit to report?

When and where do I hear about PEP and emergency contraceptives? (Are there rape kits?)

Note: If respondents mentioned special services such as rape-related services, legal

referrals or counselling services, you can ask them to show you where these are.


discussion will also explore these.

Facilitator asks respondent if (s)he has any questions for us.

Facilitator thanks respondent for her/his time and the interesting information they have

shared.

362

3. Observation and In-Depth Interview Guideline on Domestic Violence Ghana:

District Police/DOVVSU Institutions

Approach: Observations and in-depth interviews at police facilities (two per region and

>20 members of the police). In each urban and rural site, police stations are visited; in each

of these sites, both observations and interviews should take place. Purposive sample of

individuals who are directly involved in domestic violence. These can be leaders or front-

line staff equally divided across selected urban and rural sites.

Facilitator introduces the team, explains consent and the purpose.




Survey on Domestic Violence 2015 being conducted by the Institute of Development

Studies, University of Sussex (IDS) for the Ministry of Gender, Children and Social

Protection and DfiD, UK. It covers all 10 regions of Ghana, urban and rural areas.





The purpose of today’s visit and discussion is to learn from you about:

The causes and consequences of domestic violence, and which services are available

to survivors of domestic violence



The findings will inform policy recommendations to the Ministry of Gender,

Children and Social Protection. It will also enable civil society organisations to reach

out more directly to survivors of domestic violence and lobby for more

programmatic interventions and changes in laws.

There will be a report that will be published.


permission.




363





Discussion starts.

Overview, background:


Probes:

a. As a police officer, what are some of your main responsibilities?

b. As a police officer, what are some of the main challenges in your work?

i. As a police officer, do you see any problems with the way men treat

women? With the way women treat men? With the way parents treat

their children?



same family?

3. What kinds of fights do you see?

[Probe to find about forms:


b. Who is involved? [Probe for: partners, parents and children, adults and non-

biological children.]


[Probe to find out about causes. Are they about money, about infidelity, about school

etc.]








of effects – they include losing confidence, being afraid to leave the house, not being

able to work and becoming trapped in a vicious cycle of dependence, for example.

How are these effects distributed across members of the household who are affected

by the violence?]




[Probe for economic, psychological, health impacts, children’s lives.]

364







13. Discuss what are social pressures – family and community pressure to encourage or

discourage prosecution.


cases?



involved? [Probe: to find out about government and non-government service

providers:








16. What kind of services are there to help people? [Probe to get a list of services to

understand the ‘in principle’ facets of government support as the participant

understands this.]



people in their family? (Discuss satisfaction police handling of case reporting/help

at health centres.)




c. Make sure government services reach the right people to end the violence?

Which services are seen to be most crucial? Why?

Availability of logistical support

20. Does the unit have specialised departments which offer services such as:

Advocacy for victims of violence

Shelter for victims of violence

Continuing Education Units

Counselling for victims of violence

Self-help advice for victims of violence

Protection for victims of violence

Prosecution advice for victims of violence

365

Batterer’s prevention counsellors

Any gender-specific support services (name service ……………………...)

Availability of protocols

21. Does DOVVSU have formal protocols to follow when dealing with reporting of

incidences of domestic violence?

Yes

No

If Yes, name the documents …………………………………………………………

If yes, what do these protocols deal with?

Dispatch response

Arrests/service of orders

Investigation techniques

Evidence

Reports

Follow-ups

Victim’s rights

Rights perpetuators of violence

Emergency judicial response

Refuge/victim’s safety

Capacity-building

22. Have DOVVSU staff received the following specialised training to provide specific

help to the victims of violence?

Initial training on gender-based violence and on domestic violence

Continuing training on domestic violence

Domestic violence and sexual assault training

Specialised in listening skills for victim’s story

Case processing and pace of justice

23. Does DOVVSU/police station have formal procedures to follow to process domestic

violence cases?

If Yes, describe the procedure ..…………………………………………………...

………………………………………………………………………………………..

If No, how are cases of domestic violence processed?

………….…………………………………………………………………………….

How many cases have been reported in

2012

2013

2014

How many cases have been prosecuted?

2012

2013

2014

366

Continuum of care

24. Does this police facility work with any other service providers? If yes, who? (social

workers). Name them

…………....…………………………………………………………………………………..

25. How effective is this collaboration?

26. Can you give me an example of where this facility collaborated with other service

providers to give support to someone who had experienced domestic violence?

OBSERVATIONS

While waiting, look for:

Are files stored away or can the public see these easily? Do you hear staff talk about cases

and private affairs of patients in front of you when you are in a public area? Is there a

private room where victims can file a complaint?

Ask:

Can you show us the building and the services as if we are a victim of domestic violence

who comes to report? Can you show us the process step by step? Where would we go first,

and what would we do? Who would help us? Where would we sit to report?





shared.

367

4. In-Depth Interview Guideline – Domestic Violence Ghana Opinion/Community

Leaders

Approach: For in-depth interviews with opinion leaders, activists and legal experts: 20

community leaders (e.g. religious leaders, and formal and informal political leaders); 10–

20 legal practitioners; 10–20 activists (national or regional level); and 10–20 journalists

(national or regional level). Purposive sample in all regions of individuals who have been

in current position over two years, equally divided across selected urban and rural sites

alike.

Facilitator introduces the team (2 people maximum) and explains consent and the

purpose.




Survey on Domestic Violence conducted for the Ministry of Gender, Children and Social

Protection. It covers all 10 regions of Ghana, urban and rural areas. We use both

quantitative and qualitative methodologies.





The purpose of today’s discussion is to learn from you about

The causes and consequences of domestic violence, and which services survivors of

domestic violence use



The findings will also inform civil society organisations and help them to reach out

to survivors of domestic violence and lobby for more programmatic interventions

and changes in laws.

There will be a report that will be published internationally.

Facilitator explains consent forms and the value of recording, and checks permission.




368





Discussion starts.

Overview, background: tell me about yourself


[Probe:

a. As a religious/political/community leader, what are some of your main

responsibilities?

b. As a religious/political/community leader, what are some of the main

challenges in your work?

c. As a religious/political/community leader, do you see any problems with the

way men treat women? With the way women treat men? With the way

parents treat their children?]



same family?

3. What kinds of fights do you see?

[Probe to find about forms:


b. Who is involved? Probe for: partners, parents and children, adults and non-

biological children]


[Probe to find out about causes. Are they about money, about infidelity, about school

etc.?]








of effects: losing confidence, being afraid to leave the house, not being able to work

and becoming trapped in a vicious cycle of dependence, for example. How are these

effects distributed across members of the household who are affected by the

violence?]



369


[Probe for economic, psychological, health impacts.]







13. Discuss what are social pressures – family and community pressure to encourage or

discourage prosecution.


cases?



involved?

[Probe: to find out about government and non-government service providers:








16. What kind of services are there to help victims of domestic violence? (Get a list of

services to understand the ‘in principle’ facets of government support as the

participant understands this.)



people in their family? (Discuss here satisfaction with police handling of case

reporting/help at health centres.) Are there interventions to prevent domestic

violence in this community?

.……………………………………………………………………………………………….

19. Which of these interventions work against domestic violence? Why?

20. What can the government do better, or differently, to make sure that its policy is

properly implemented?

a. In rural areas?

b. In peri-urban areas?

c. In urban areas?

d. Through: Police? Health workers? Social workers?




370

c. Help government services reach the right people to end the violence?

d. Which services are seen to be most crucial? Why?





shared.

371

5. In-Depth Interview Guideline – Victims/Survivors of Domestic Violence Ghana

Approach: For in-depth interviews with 45 survivors of domestic violence (three per

enumerator) identified during the quantitative research – the survey. Purposive sample in

all regions of individuals who self-identify as victims of domestic violence, equally divided

across urban and rural areas alike.

Facilitator introduces him or herself, explains consent and the purpose.



My colleague [name of note-taker], will take notes. This study is being conducted for the

Ministry of Gender, Children and Social Protection. It covers all 10 regions of Ghana, urban

and rural areas.

Your participation in this research is voluntary. The discussion may last for about 30 to 60




You were invited for this interview because you participated in a survey during which you

self-identified as someone who has or is experiencing domestic violence. Thank you for

agreeing to speak some more with us about your experiences.

The purpose of today’s discussion is to learn from you about:

the causes and consequences of domestic violence;

which services are available to victims/survivors of domestic violence;

what works well for victims/survivors of domestic violence, and what does not; and

how these services can be improved or strengthened.

There will be a report that will be published from this country-wide study; however,

individual stories will not be identified, so no one will know that you spoke to us or what

you told us.

Facilitator explains about consent forms and the value of recording, and checks

permission. Facilitator helps to complete the consent form.

Discussion starts.

Background information

1. Can you tell me a little bit about yourself? [Interviewer must ask and record the

following background information for each interview: age; sex; marital status

(single/married/living together/living together with children); educational level;

region; district; city/town/rural site of interview.]

372

2. About your home, and who you live with?

3. About your family, who forms part of the house/household (nuclear or extended family

compound household/ rented accommodation etc.); and who do you take care of/takes

care of you in the home?

4. About your daily activities, and what takes up your time during the day (earning a

living or supplementing income/household chores or combination of these etc.)?

5. About growing up, where you grew up and how you came to live here?


6. You told us that you experienced some violence from other family members of your

household in your life? Is that correct?

7. Can you tell me a little bit more about this experience? [Interviewer should give the

respondent the opportunity to briefly narrate uninterrupted their experience of

domestic violence – notes should be taken of this narration.]

8. Is this situation of violence ongoing? Or was it in the past? When?

9. Let us go into more detail here. What kinds of fights and disagreement did/do you

experience/see in your own family?

[Probe to find about forms: Are these fights verbal, physical, sexual, economic or

emotional/psychological?]

10. Who was/are involved?

[Probe to find out about domestic relations: Partners, parents and children, siblings,

aunts, uncles, cousins, grandparents, stepchildren and non-biological children,

house help.]

11. Why does this fighting/abuse and violence take place in your family?

[Probe to find out about causes: Are they about money, about infidelity, about school

etc.?]

12. How often does this fighting and violence take place?

[Probe for incidence: At least once a week or once a month? Are there specific times

when this fighting takes place? Any seasons or specific days? Is it linked to school

terms? Is it linked to monthly payment cycles?]


13. Who is affected by fights in your family? How? Why these people?

[Probe for the range of effects – they may include:

a. Victim losing confidence, being afraid to leave the house, not being able to

work and becoming trapped in a vicious cycle of dependence.

b. The effects distributed across members of the household who are affected by

the violence (by age, sex, biological children/foster children/other kin).

c. How are they impacted by the incidences of violence?]


[Probe for short-/long-term impacts: economic, psychological, health impacts.]

Victim responses to violence

15. When you and your family have conflicts, what do you do to cope with the violence?

373

16. When a person is angry or uses violence or threatens to use violence against you, have

you ever sought support from other people in the family?


17. Does anyone in the family come to intervene/help you in any situations of family

violence? Who? How? Why?

18. Is there any one to help reconcile the conflicts between you and the person who has

hurt you in the family? Please explain.

[Probe for family/extended family intervention and mediation.]

19. Have you been helped by neighbours or local authorities/local organisations (i.e. chief

or queen mother, women’s union, reconciliation group)? Please explain the reasons

why you have or have not sought help from these people/organisations.

[Probe: Community intervention and mediation.

a. If no, why did you not seek help from these people community facilities?

b. If yes, who/which did you go to for assistance? What were their responses

when you told them your story or asked them for help? How did these

help/support you? Were you satisfied with the help they provided? Did their

support help end the violence you were facing from the family member?]

Knowledge/availability of victim support services

20. Can you tell us what you know to do/have been advised to do when you are hurt by

members of your family?

[Probe for knowledge of rights/resources:

a. Do you know about the Ghana 2007 Domestic Violence Act 732/gender

equity laws?

b. Do you know about any domestic violence victim support services (ask

specifically about DOVVSU, CHRAJ, others)?]

21. When you were injured from the violence, did you go to the police for help?

[Probe for use/availability of support services and programmes:

a. If no, why did you not use the police facilities?

b. If yes, how did the police facility help/support you?

c. Were you satisfied with the help they provided?

d. Did the police support help end the violence you were facing from the family

member?]

22. When you were injured from the violence, did you go to health facilities such as

hospital, commune health station, pharmacy stores?

[Probe:

a. If no, why did you not use these facilities?

b. If yes, which of the health facilities did you go to (public or private)?

c. How did the health providers help/support you?

d. Did you have to pay for the services?

e. If yes, did you have enough money to pay for the services?

f. Did you have someone to accompany you, or did you go by yourself? If yes,

who?]

374

23. What do you think local leaders can do to prevent/support victims of domestic violence

(to find out about the role of chiefs, community leaders, local assembly/MP and other

administrative/political officials play in implementing policy directives)?

24. What do you think the police facilities can do to support victims of domestic violence

(to find out about police support and help to prevent domestic violence)?

25. What do you think the health facilities can do to support victims of domestic violence

(to find out about health support for victims and help to prevent domestic violence)?

26. What do you think are the most effective ways to stop other people from being hurt

like this?

[Probe:

a. Are these types of support available? Where do you think you can find them?

b. How does/did the absence of any of these types of support influence your

situation?]

27. Looking back at your situation, what advice would you give another woman or man,

or child, who has just started to have these sorts of problems in their home?

[Probe: What should these people do? Ask about all the options/choices they have or

do not have]

28. How do you think people who hurt other people, like you have been hurt, should be

punished? What should happen?

[Probe: How do people think about justice, and appropriate punishment?]

29. Are you a member of a group or an organisation? Which (religious, social or traditional

group, saving and credit or any group)?

30. If you could advise the government, what would you tell it to do? [This is an open-

ended question that can be really useful in drawing out unique responses about

government accountability to end domestic violence, so encourage respondent.]

N.B. These are guidelines. If any interesting additional issues emerge, the discussion

should also explore these.

WRAP UP

Facilitator asks respondent if (s)he has any questions for interviewing team.

Facilitator thanks respondent for her/his time. (Thank you for sharing this with us. We

have asked very difficult questions, and thank you for being so open. All shared

information will remain confidential. What you have told us is very important, and will

help us in our work to address domestic violence.)

From the victim’s responses, mention his/her strengths.

Give details of follow-up counselling support available both immediately and later.

Give more general information about services available in the community.

:: ghana statistical service :: - reflect the uk governments official … · 2016. 11. 21. · this...

Documents